*
* autovacuum.cpp
*
* openGauss Integrated Autovacuum Daemon
*
* The autovacuum system is structured in two different kinds of processes: the
* autovacuum launcher and the autovacuum worker. The launcher is an
* always-running process, started by the postmaster when the autovacuum GUC
* parameter is set. The launcher schedules autovacuum workers to be started
* when appropriate. The workers are the processes which execute the actual
* vacuuming; they connect to a database as determined in the launcher, and
* once connected they examine the catalogs to select the tables to vacuum.
*
* The autovacuum launcher cannot start the worker processes by itself,
* because doing so would cause robustness issues (namely, failure to shut
* them down on exceptional conditions, and also, since the launcher is
* connected to shared memory and is thus subject to corruption there, it is
* not as robust as the postmaster). So it leaves that task to the postmaster.
*
* There is an autovacuum shared memory area, where the launcher stores
* information about the database it wants vacuumed. When it wants a new
* worker to start, it sets a flag in shared memory and sends a signal to the
* postmaster. Then postmaster knows nothing more than it must start a worker;
* so it forks a new child, which turns into a worker. This new process
* connects to shared memory, and there it can inspect the information that the
* launcher has set up.
*
* If the fork() call fails in the postmaster, it sets a flag in the shared
* memory area, and sends a signal to the launcher. The launcher, upon
* noticing the flag, can try starting the worker again by resending the
* signal. Note that the failure can only be transient (fork failure due to
* high load, memory pressure, too many processes, etc); more permanent
* problems, like failure to connect to a database, are detected later in the
* worker and dealt with just by having the worker exit normally. The launcher
* will launch a new worker again later, per schedule.
*
* When the worker is done vacuuming it sends SIGUSR2 to the launcher. The
* launcher then wakes up and is able to launch another worker, if the schedule
* is so tight that a new worker is needed immediately. At this time the
* launcher can also balance the settings for the various remaining workers'
* cost-based vacuum delay feature.
*
* Note that there can be more than one worker in a database concurrently.
* They will store the table they are currently vacuuming in shared memory, so
* that other workers avoid being blocked waiting for the vacuum lock for that
* table. They will also reload the pgstats data just before vacuuming each
* table, to avoid vacuuming a table that was just finished being vacuumed by
* another worker and thus is no longer noted in shared memory. However,
* there is a window (caused by pgstat delay) on which a worker may choose a
* table that was already vacuumed; this is a bug in the current design.
*
* Portions Copyright (c) 2020 Huawei Technologies Co.,Ltd.
* Portions Copyright (c) 1996-2012, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
* Portions Copyright (c) 2021, openGauss Contributors
*
*
* IDENTIFICATION
* src/gausskernel/process/postmaster/autovacuum.cpp
*
* -------------------------------------------------------------------------
*/
#include "postgres.h"
#include "knl/knl_variable.h"
#include <signal.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/time.h>
#include <time.h>
#include <unistd.h>
#ifdef HAVE_SYS_STAT_H
#include <sys/stat.h>
#endif
#include "access/hash.h"
#include "access/gtm.h"
#include "access/heapam.h"
#include "access/reloptions.h"
#include "access/tableam.h"
#include "access/twophase.h"
#include "access/transam.h"
#include "access/xact.h"
#include "access/multixact.h"
#include "catalog/dependency.h"
#include "catalog/namespace.h"
#include "catalog/pg_database.h"
#include "catalog/pgxc_class.h"
#include "catalog/pg_statistic_history.h"
#include "catalog/pg_statistic_lock.h"
#include "commands/dbcommands.h"
#include "commands/vacuum.h"
#include "commands/matview.h"
#include "executor/executor.h"
#include "executor/spi.h"
#include "funcapi.h"
#include "libpq/pqsignal.h"
#include "miscadmin.h"
#include "pgstat.h"
#include "pgxc/execRemote.h"
#include "pgxc/pgxc.h"
#include "postmaster/autovacuum.h"
#include "postmaster/fork_process.h"
#include "postmaster/postmaster.h"
#include "storage/buf/bufmgr.h"
#include "storage/ipc.h"
#include "storage/latch.h"
#include "storage/pmsignal.h"
#include "storage/proc.h"
#include "storage/procarray.h"
#include "storage/procsignal.h"
#include "storage/sinvaladt.h"
#include "tcop/tcopprot.h"
#include "utils/fmgroids.h"
#include "utils/lsyscache.h"
#include "utils/memutils.h"
#include "utils/postinit.h"
#include "utils/ps_status.h"
#include "utils/rel.h"
#include "utils/rel_gs.h"
#include "utils/snapmgr.h"
#include "utils/syscache.h"
#include "utils/timestamp.h"
#include "access/heapam.h"
#include "commands/user.h"
#include "gssignal/gs_signal.h"
#include "common/config/cm_config.h"
#include "catalog/pg_namespace.h"
#include "storage/lmgr.h"
typedef struct avw_info {
PgStat_StatTabKey tabkey;
int64 n_dead_tuples;
int64 n_live_tuples;
int64 changes_since_analyze;
} avw_info;
typedef struct avw_dbase {
Oid adw_datid;
char* adw_name;
TransactionId adw_frozenxid;
MultiXactId adw_frozenmulti;
PgStat_StatDBEntry* adw_entry;
} avw_dbase;
NON_EXEC_STATIC void AutoVacWorkerMain();
NON_EXEC_STATIC void AutoVacLauncherMain();
static Oid do_start_worker(void);
static void launcher_determine_sleep(bool canlaunch, bool recursing, struct timeval* nap);
static void launch_worker(TimestampTz now);
static List* get_database_list(void);
static void rebuild_database_list(Oid newdb);
static int db_comparator(const void* a, const void* b);
static void autovac_balance_cost(void);
static void do_autovacuum(void);
static void FreeWorkerInfo(int code, Datum arg);
static autovac_table* table_recheck_autovac(
vacuum_object* vacObj, HTAB* table_toast_map, HTAB* toast_table_map, TupleDesc pg_class_desc);
static void relation_needs_vacanalyze(Oid relid, AutoVacOpts* relopts, bytea* rawRelopts, Form_pg_class classForm,
HeapTuple tuple, PgStat_StatTabEntry* tabentry, bool allowAnalyze, bool allowVacuum, bool is_recheck,
bool* dovacuum, bool* doanalyze, bool* need_freeze);
static void autovacuum_do_vac_analyze(autovac_table* tab, BufferAccessStrategy bstrategy);
static void autovacuum_local_vac_analyze(autovac_table* tab, BufferAccessStrategy bstrategy);
static PgStat_StatTabEntry* get_pgstat_tabentry_relid(
Oid relid, bool isshared, uint32 statFlag, PgStat_StatDBEntry* shared, PgStat_StatDBEntry* dbentry);
static void autovac_report_activity(autovac_table* tab);
static void avl_sighup_handler(SIGNAL_ARGS);
static void avl_sigusr2_handler(SIGNAL_ARGS);
static void avl_sigterm_handler(SIGNAL_ARGS);
static void autovac_refresh_stats(void);
static void partition_needs_vacanalyze(Oid partid, AutoVacOpts* relopts, Form_pg_partition partForm,
HeapTuple partTuple, at_partitioned_table* ap_entry, PgStat_StatTabEntry* tabentry, bool is_recheck, bool* dovacuum,
bool* doanalyze, bool* need_freeze);
static autovac_table* partition_recheck_autovac(
vacuum_object* vacObj, HTAB* table_relopt_map, HTAB* partitioned_tables_map, TupleDesc pg_class_desc);
extern void DoVacuumMppTable(VacuumStmt* stmt, const char* queryString, bool isTopLevel, bool sentToRemote);
* Called when the AutoVacuum is ending.
*/
static void autoVacQuitAndClean(int code, Datum arg)
{
CloseGTM();
free_RemoteXactState();
UnsetGlobalSnapshotData();
}
* AUTOVACUUM LAUNCHER CODE
********************************************************************/
#ifdef EXEC_BACKEND
* We need this set from the outside, before InitProcess is called
*/
void AutovacuumLauncherIAm(void)
{
t_thrd.role = AUTOVACUUM_LAUNCHER;
}
#endif
* Main loop for the autovacuum launcher process.
*/
NON_EXEC_STATIC void AutoVacLauncherMain()
{
sigjmp_buf local_sigjmp_buf;
IsUnderPostmaster = true;
t_thrd.role = AUTOVACUUM_LAUNCHER;
t_thrd.proc_cxt.MyProcPid = gs_thread_self();
t_thrd.proc_cxt.MyStartTime = time(NULL);
t_thrd.proc_cxt.MyProgName = "AutoVacLauncher";
u_sess->attr.attr_common.application_name = pstrdup("AutoVacLauncher");
init_ps_display("autovacuum launcher process", "", "", "");
ereport(LOG, (errmsg("autovacuum launcher started")));
if (u_sess->attr.attr_security.PostAuthDelay)
pg_usleep(u_sess->attr.attr_security.PostAuthDelay * 1000000L);
SetProcessingMode(InitProcessing);
* Set up signal handlers. We operate on databases much like a regular
* backend, so we use the same signal handling. See equivalent code in
* tcop/postgres.c.
*/
gspqsignal(SIGHUP, avl_sighup_handler);
gspqsignal(SIGINT, StatementCancelHandler);
gspqsignal(SIGTERM, avl_sigterm_handler);
gspqsignal(SIGQUIT, quickdie);
gspqsignal(SIGALRM, handle_sig_alarm);
gspqsignal(SIGPIPE, SIG_IGN);
gspqsignal(SIGUSR1, procsignal_sigusr1_handler);
gspqsignal(SIGUSR2, avl_sigusr2_handler);
gspqsignal(SIGFPE, FloatExceptionHandler);
gspqsignal(SIGCHLD, SIG_DFL);
gspqsignal(SIGURG, print_stack);
BaseInit();
* Create a per-backend PGPROC struct in shared memory, except in the
* EXEC_BACKEND case where this was done in SubPostmasterMain. We must do
* this before we can use LWLocks (and in the EXEC_BACKEND case we already
* had to do some stuff with LWLocks).
*/
#ifndef EXEC_BACKEND
InitProcess();
#endif
t_thrd.proc_cxt.PostInit->SetDatabaseAndUser(NULL, InvalidOid, NULL);
t_thrd.proc_cxt.PostInit->InitAutoVacLauncher();
SetProcessingMode(NormalProcessing);
on_proc_exit(autoVacQuitAndClean, 0);
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks.
*/
t_thrd.autovacuum_cxt.AutovacMemCxt = AllocSetContextCreate(t_thrd.top_mem_cxt, "Autovacuum Launcher",
ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MAXSIZE);
(void)MemoryContextSwitchTo(t_thrd.autovacuum_cxt.AutovacMemCxt);
* If an exception is encountered, processing resumes here.
*
* This code is a stripped down version of PostgresMain error recovery.
*/
int curTryCounter;
int* oldTryCounter = NULL;
if (sigsetjmp(local_sigjmp_buf, 1) != 0) {
gstrace_tryblock_exit(true, oldTryCounter);
t_thrd.log_cxt.error_context_stack = NULL;
t_thrd.log_cxt.call_stack = NULL;
HOLD_INTERRUPTS();
t_thrd.int_cxt.QueryCancelPending = false;
disable_sig_alarm(true);
t_thrd.int_cxt.QueryCancelPending = false;
EmitErrorReport();
AbortCurrentTransaction();
AtEOXact_SysDBCache(false);
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
(void)MemoryContextSwitchTo(t_thrd.autovacuum_cxt.AutovacMemCxt);
FlushErrorState();
MemoryContextResetAndDeleteChildren(t_thrd.autovacuum_cxt.AutovacMemCxt);
t_thrd.autovacuum_cxt.DatabaseListCxt = NULL;
t_thrd.autovacuum_cxt.DatabaseList = NULL;
* Make sure pgstat also considers our stat data as gone. Note: we
* mustn't use autovac_refresh_stats here.
*/
pgstat_clear_snapshot();
RESUME_INTERRUPTS();
if (t_thrd.autovacuum_cxt.got_SIGTERM)
goto shutdown;
* Sleep at least 1 second after any error. We don't want to be
* filling the error logs as fast as we can.
*/
pg_usleep(1000000L);
}
oldTryCounter = gstrace_tryblock_entry(&curTryCounter);
t_thrd.log_cxt.PG_exception_stack = &local_sigjmp_buf;
gs_signal_setmask(&t_thrd.libpq_cxt.UnBlockSig, NULL);
gs_signal_unblock_sigusr2();
* Force zero_damaged_pages OFF in the autovac process, even if it is set
* in postgresql.conf. We don't really want such a dangerous option being
* applied non-interactively.
*/
SetConfigOption("zero_damaged_pages", "false", PGC_SUSET, PGC_S_OVERRIDE);
* Force statement_timeout to zero to avoid a timeout setting from
* preventing regular maintenance from being executed.
*/
SetConfigOption("statement_timeout", "0", PGC_SUSET, PGC_S_OVERRIDE);
* Force default_transaction_isolation to READ COMMITTED. We don't want
* to pay the overhead of serializable mode, nor add any risk of causing
* deadlocks or delaying other transactions.
*/
SetConfigOption("default_transaction_isolation", "read committed", PGC_SUSET, PGC_S_OVERRIDE);
* In emergency mode, just start a worker (unless shutdown was requested)
* and go away.
*/
if (!AutoVacuumingActive()) {
if (!t_thrd.autovacuum_cxt.got_SIGTERM)
do_start_worker();
proc_exit(0);
}
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_launcherpid = t_thrd.proc_cxt.MyProcPid;
if (likely(t_thrd.shemem_ptr_cxt.mySessionMemoryEntry != NULL)) {
t_thrd.shemem_ptr_cxt.mySessionMemoryEntry->initMemInChunks = t_thrd.utils_cxt.trackedMemChunks;
t_thrd.shemem_ptr_cxt.mySessionMemoryEntry->queryMemInChunks = t_thrd.utils_cxt.trackedMemChunks;
t_thrd.shemem_ptr_cxt.mySessionMemoryEntry->peakChunksQuery = t_thrd.utils_cxt.trackedMemChunks;
}
t_thrd.utils_cxt.peakedBytesInQueryLifeCycle = 0;
t_thrd.utils_cxt.basedBytesInQueryLifeCycle = 0;
* Create the initial database list. The invariant we want this list to
* keep is that it's ordered by decreasing next_time. As soon as an entry
* is updated to a higher time, it will be moved to the front (which is
* correct because the only operation is to add autovacuum_naptime to the
* entry, and time always increases).
*/
rebuild_database_list(InvalidOid);
while (!t_thrd.autovacuum_cxt.got_SIGTERM) {
struct timeval nap;
TimestampTz current_time = 0;
bool can_launch = false;
Dlelem* elem = NULL;
int rc;
* This loop is a bit different from the normal use of WaitLatch,
* because we'd like to sleep before the first launch of a child
* process. So it's WaitLatch, then ResetLatch, then check for
* wakening conditions.
*/
launcher_determine_sleep((t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers != NULL), false, &nap);
* Wait until naptime expires or we get some type of signal (all the
* signal handlers will wake us by calling SetLatch).
*/
rc = WaitLatch(&t_thrd.proc->procLatch, WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
(nap.tv_sec * 1000L) + (nap.tv_usec / 1000L));
ResetLatch(&t_thrd.proc->procLatch);
ProcessCatchupInterrupt();
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (((unsigned int)rc) & WL_POSTMASTER_DEATH)
proc_exit(1);
if (t_thrd.autovacuum_cxt.got_SIGTERM)
break;
if (t_thrd.autovacuum_cxt.got_SIGHUP) {
t_thrd.autovacuum_cxt.got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
if (!AutoVacuumingActive())
break;
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
autovac_balance_cost();
LWLockRelease(AutovacuumLock);
rebuild_database_list(InvalidOid);
}
* a worker finished, or postmaster signalled failure to start a
* worker
*/
if (t_thrd.autovacuum_cxt.got_SIGUSR2) {
t_thrd.autovacuum_cxt.got_SIGUSR2 = false;
if (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_signal[AutoVacRebalance]) {
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_signal[AutoVacRebalance] = false;
autovac_balance_cost();
LWLockRelease(AutovacuumLock);
}
if (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_signal[AutoVacForkFailed]) {
* If the postmaster failed to start a new worker, we sleep
* for a little while and resend the signal. The new worker's
* state is still in memory, so this is sufficient. After
* that, we restart the main loop.
*
* XXX should we put a limit to the number of times we retry?
* I don't think it makes much sense, because a future start
* of a worker will continue to fail in the same way.
*/
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_signal[AutoVacForkFailed] = false;
pg_usleep(1000000L);
SendPostmasterSignal(PMSIGNAL_START_AUTOVAC_WORKER);
continue;
}
}
* There are some conditions that we need to check before trying to
* start a launcher. First, we need to make sure that there is a
* launcher slot available. Second, we need to make sure that no
* other worker failed while starting up.
*/
current_time = GetCurrentTimestamp();
LWLockAcquire(AutovacuumLock, LW_SHARED);
can_launch = (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers != NULL);
if (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker != NULL) {
int waittime;
WorkerInfo worker = t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker;
* We can't launch another worker when another one is still
* starting up (or failed while doing so), so just sleep for a bit
* more; that worker will wake us up again as soon as it's ready.
* We will only wait autovacuum_naptime seconds (up to a maximum
* of 60 seconds) for this to happen however. Note that failure
* to connect to a particular database is not a problem here,
* because the worker removes itself from the startingWorker
* pointer before trying to connect. Problems detected by the
* postmaster (like fork() failure) are also reported and handled
* differently. The only problems that may cause this code to
* fire are errors in the earlier sections of AutoVacWorkerMain,
* before the worker removes the WorkerInfo from the
* startingWorker pointer.
*/
waittime = Min(u_sess->attr.attr_storage.autovacuum_naptime, 60) * 1000;
if (TimestampDifferenceExceeds(worker->wi_launchtime, current_time, waittime)) {
LWLockRelease(AutovacuumLock);
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
* No other process can put a worker in starting mode, so if
* startingWorker is still INVALID after exchanging our lock,
* we assume it's the same one we saw above (so we don't
* recheck the launch time).
*/
if (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker != NULL) {
worker = t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker;
worker->wi_dboid = InvalidOid;
worker->wi_tableoid = InvalidOid;
worker->wi_parentoid = InvalidOid;
worker->wi_sharedrel = false;
worker->wi_proc = NULL;
worker->wi_launchtime = 0;
worker->wi_links.next = (SHM_QUEUE*)t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers;
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers = worker;
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker = NULL;
ereport(WARNING, (errmsg("worker took too long to start; canceled")));
}
} else
can_launch = false;
}
LWLockRelease(AutovacuumLock);
if (!can_launch)
continue;
elem = DLGetTail(t_thrd.autovacuum_cxt.DatabaseList);
if (elem != NULL) {
avl_dbase* avdb = (avl_dbase*)DLE_VAL(elem);
* launch a worker if next_worker is right now or it is in the
* past
*/
if (TimestampDifferenceExceeds(avdb->adl_next_worker, current_time, 0))
launch_worker(current_time);
} else {
* Special case when the list is empty: start a worker right away.
* This covers the initial case, when no database is in pgstats
* (thus the list is empty). Note that the constraints in
* launcher_determine_sleep keep us from starting workers too
* quickly (at most once every autovacuum_naptime when the list is
* empty).
*/
launch_worker(current_time);
}
}
shutdown:
ereport(LOG, (errmsg("autovacuum launcher shutting down")));
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_launcherpid = 0;
proc_exit(0);
}
* Determine the time to sleep, based on the database list.
*
* The "canlaunch" parameter indicates whether we can start a worker right now,
* for example due to the workers being all busy. If this is false, we will
* cause a long sleep, which will be interrupted when a worker exits.
*/
static void launcher_determine_sleep(bool canlaunch, bool recursing, struct timeval* nap)
{
Dlelem* elem = NULL;
* We sleep until the next scheduled vacuum. We trust that when the
* database list was built, care was taken so that no entries have times
* in the past; if the first entry has too close a next_worker value, or a
* time in the past, we will sleep a small nominal time.
*/
if (!canlaunch) {
nap->tv_sec = u_sess->attr.attr_storage.autovacuum_naptime;
nap->tv_usec = 0;
} else if ((elem = DLGetTail(t_thrd.autovacuum_cxt.DatabaseList)) != NULL) {
avl_dbase* avdb = (avl_dbase*)DLE_VAL(elem);
TimestampTz current_time = GetCurrentTimestamp();
TimestampTz next_wakeup;
long secs;
int usecs;
next_wakeup = avdb->adl_next_worker;
TimestampDifference(current_time, next_wakeup, &secs, &usecs);
nap->tv_sec = secs;
nap->tv_usec = usecs;
} else {
nap->tv_sec = u_sess->attr.attr_storage.autovacuum_naptime;
nap->tv_usec = 0;
}
* If the result is exactly zero, it means a database had an entry with
* time in the past. Rebuild the list so that the databases are evenly
* distributed again, and recalculate the time to sleep. This can happen
* if there are more tables needing vacuum than workers, and they all take
* longer to vacuum than autovacuum_naptime.
*
* We only recurse once. rebuild_database_list should always return times
* in the future, but it seems best not to trust too much on that.
*/
if (nap->tv_sec == 0 && nap->tv_usec == 0 && !recursing) {
rebuild_database_list(InvalidOid);
launcher_determine_sleep(canlaunch, true, nap);
return;
}
if (nap->tv_sec <= 0 && nap->tv_usec <= MIN_AUTOVAC_SLEEPTIME * 1000) {
nap->tv_sec = 0;
nap->tv_usec = (__suseconds_t)(MIN_AUTOVAC_SLEEPTIME * 1000);
}
}
* Build an updated t_thrd.autovacuum_cxt.DatabaseList. It must only contain databases that appear
* in pgstats, and must be sorted by next_worker from highest to lowest,
* distributed regularly across the next autovacuum_naptime interval.
*
* Receives the Oid of the database that made this list be generated (we call
* this the "new" database, because when the database was already present on
* the list, we expect that this function is not called at all). The
* preexisting list, if any, will be used to preserve the order of the
* databases in the autovacuum_naptime period. The new database is put at the
* end of the interval. The actual values are not saved, which should not be
* much of a problem.
*/
static void rebuild_database_list(Oid newdb)
{
List* dblist = NIL;
ListCell* cell = NULL;
MemoryContext newcxt;
MemoryContext oldcxt;
MemoryContext tmpcxt;
HASHCTL hctl;
uint score;
uint nelems;
HTAB* dbhash = NULL;
autovac_refresh_stats();
newcxt = AllocSetContextCreate(t_thrd.autovacuum_cxt.AutovacMemCxt, "AV dblist",
ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MAXSIZE);
tmpcxt = AllocSetContextCreate(
newcxt, "tmp AV dblist", ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MAXSIZE);
oldcxt = MemoryContextSwitchTo(tmpcxt);
* Implementing this is not as simple as it sounds, because we need to put
* the new database at the end of the list; next the databases that were
* already on the list, and finally (at the tail of the list) all the
* other databases that are not on the existing list.
*
* To do this, we build an empty hash table of scored databases. We will
* start with the lowest score (zero) for the new database, then
* increasing scores for the databases in the existing list, in order, and
* lastly increasing scores for all databases gotten via
* get_database_list() that are not already on the hash.
*
* Then we will put all the hash elements into an array, sort the array by
* score, and finally put the array elements into the new doubly linked
* list.
*/
hctl.keysize = sizeof(Oid);
hctl.entrysize = sizeof(avl_dbase);
hctl.hash = oid_hash;
hctl.hcxt = tmpcxt;
dbhash = hash_create("db hash", 20, &hctl, HASH_ELEM | HASH_FUNCTION | HASH_CONTEXT);
score = 0;
if (OidIsValid(newdb)) {
avl_dbase* db = NULL;
PgStat_StatDBEntry* entry = NULL;
entry = pgstat_fetch_stat_dbentry(newdb);
if (entry != NULL) {
db = (avl_dbase*)hash_search(dbhash, &newdb, HASH_ENTER, NULL);
db->adl_score = score++;
}
}
if (t_thrd.autovacuum_cxt.DatabaseList != NULL) {
Dlelem* elem = NULL;
elem = DLGetHead(t_thrd.autovacuum_cxt.DatabaseList);
while (elem != NULL) {
avl_dbase* avdb = (avl_dbase*)DLE_VAL(elem);
avl_dbase* db = NULL;
bool found = false;
PgStat_StatDBEntry* entry = NULL;
elem = DLGetSucc(elem);
* skip databases with no stat entries -- in particular, this gets
* rid of dropped databases
*/
entry = pgstat_fetch_stat_dbentry(avdb->adl_datid);
if (entry == NULL)
continue;
db = (avl_dbase*)hash_search(dbhash, &(avdb->adl_datid), HASH_ENTER, &found);
if (!found) {
db->adl_score = score++;
}
}
}
dblist = get_database_list();
foreach (cell, dblist) {
avw_dbase* avdb = (avw_dbase*)lfirst(cell);
avl_dbase* db = NULL;
bool found = false;
PgStat_StatDBEntry* entry = NULL;
entry = pgstat_fetch_stat_dbentry(avdb->adw_datid);
if (entry == NULL)
continue;
db = (avl_dbase*)hash_search(dbhash, &(avdb->adw_datid), HASH_ENTER, &found);
if (!found) {
db->adl_score = score++;
}
}
nelems = score;
(void)MemoryContextSwitchTo(newcxt);
t_thrd.autovacuum_cxt.DatabaseList = DLNewList();
if (nelems > 0) {
TimestampTz current_time;
int millis_increment;
avl_dbase* dbary = NULL;
avl_dbase* db = NULL;
HASH_SEQ_STATUS seq;
uint i;
int rc = 0;
if (unlikely(nelems > MaxAllocSize / sizeof(avl_dbase))) {
ereport(ERROR, (errcode(ERRCODE_INVALID_PARAMETER_VALUE), errmsg("Invalid database num:%u", nelems)));
}
dbary = (avl_dbase*)palloc(nelems * sizeof(avl_dbase));
i = 0;
hash_seq_init(&seq, dbhash);
while ((db = (avl_dbase*)hash_seq_search(&seq)) != NULL) {
rc = memcpy_s(&(dbary[i++]), sizeof(avl_dbase), db, sizeof(avl_dbase));
securec_check(rc, "\0", "\0");
}
qsort(dbary, nelems, sizeof(avl_dbase), db_comparator);
* Determine the time interval between databases in the schedule. If
* we see that the configured naptime would take us to sleep times
* lower than our min sleep time (which launcher_determine_sleep is
* coded not to allow), silently use a larger naptime (but don't touch
* the GUC variable).
*/
millis_increment = (int)(1000.0 * u_sess->attr.attr_storage.autovacuum_naptime / nelems);
if (millis_increment <= MIN_AUTOVAC_SLEEPTIME) {
millis_increment = (int)(MIN_AUTOVAC_SLEEPTIME * 1.1);
}
current_time = GetCurrentTimestamp();
* move the elements from the array into the dllist, setting the
* next_worker while walking the array
*/
for (i = 0; i < nelems; i++) {
avl_dbase* db_tmp = (avl_dbase*)&(dbary[i]);
Dlelem* elem = NULL;
current_time = TimestampTzPlusMilliseconds(current_time, millis_increment);
db_tmp->adl_next_worker = current_time;
elem = DLNewElem(db_tmp);
DLAddHead(t_thrd.autovacuum_cxt.DatabaseList, elem);
}
}
if (t_thrd.autovacuum_cxt.DatabaseListCxt != NULL)
MemoryContextDelete(t_thrd.autovacuum_cxt.DatabaseListCxt);
MemoryContextDelete(tmpcxt);
t_thrd.autovacuum_cxt.DatabaseListCxt = newcxt;
(void)MemoryContextSwitchTo(oldcxt);
}
static int db_comparator(const void* a, const void* b)
{
if (((const avl_dbase*)a)->adl_score == ((const avl_dbase*)b)->adl_score) {
return 0;
} else {
return (((const avl_dbase*)a)->adl_score < ((const avl_dbase*)b)->adl_score) ? 1 : -1;
}
}
* do_start_worker
*
* Bare-bones procedure for starting an autovacuum worker from the launcher.
* It determines what database to work on, sets up shared memory stuff and
* signals postmaster to start the worker. It fails gracefully if invoked when
* autovacuum_workers are already active.
*
* Return value is the OID of the database that the worker is going to process,
* or InvalidOid if no worker was actually started.
*/
static Oid do_start_worker(void)
{
List* dblist = NIL;
ListCell* cell = NULL;
TransactionId xidForceLimit;
#ifndef ENABLE_MULTIPLE_NODES
MultiXactId multiForceLimit;
#endif
bool for_xid_wrap = false;
bool for_multi_wrap = false;
avw_dbase* avdb = NULL;
TimestampTz current_time;
bool skipit = false;
Oid retval = InvalidOid;
MemoryContext tmpcxt, oldcxt;
LWLockAcquire(AutovacuumLock, LW_SHARED);
if (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers == NULL) {
LWLockRelease(AutovacuumLock);
AUTOVAC_LOG(LOG, "no free autovaccm worker");
return InvalidOid;
}
LWLockRelease(AutovacuumLock);
* Create and switch to a temporary context to avoid leaking the memory
* allocated for the database list.
*/
tmpcxt = AllocSetContextCreate(CurrentMemoryContext, "Start worker tmp cxt",
ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MAXSIZE);
oldcxt = MemoryContextSwitchTo(tmpcxt);
autovac_refresh_stats();
dblist = get_database_list();
* Determine the oldest datfrozenxid64/relfrozenxid64 that we will allow to
* pass without forcing a vacuum. (This limit can be tightened for
* particular tables, but not loosened.)
*/
t_thrd.autovacuum_cxt.recentXid = ReadNewTransactionId();
if (t_thrd.autovacuum_cxt.recentXid >
FirstNormalTransactionId + (uint64)g_instance.attr.attr_storage.autovacuum_freeze_max_age)
xidForceLimit = t_thrd.autovacuum_cxt.recentXid -
(uint64)g_instance.attr.attr_storage.autovacuum_freeze_max_age;
else
xidForceLimit = FirstNormalTransactionId;
#ifndef ENABLE_MULTIPLE_NODES
t_thrd.autovacuum_cxt.recentMulti = ReadNextMultiXactId();
if (t_thrd.autovacuum_cxt.recentMulti >
FirstMultiXactId + (uint64)g_instance.attr.attr_storage.autovacuum_freeze_max_age)
multiForceLimit = t_thrd.autovacuum_cxt.recentMulti -
(uint64)g_instance.attr.attr_storage.autovacuum_freeze_max_age;
else
multiForceLimit = FirstMultiXactId;
#endif
* Choose a database to connect to. We pick the database that was least
* recently auto-vacuumed, or one that needs vacuuming to recycle clog.
*
* Note that a database with no stats entry is not considered. The theory
* is that if no one has ever connected to it since the stats were last
* initialized, it doesn't need vacuuming.
*
* XXX This could be improved if we had more info about whether it needs
* vacuuming before connecting to it. Perhaps look through the pgstats
* data for the database's tables? One idea is to keep track of the
* number of new and dead tuples per database in pgstats. However it
* isn't clear how to construct a metric that measures that and not cause
* starvation for less busy databases.
*/
avdb = NULL;
for_xid_wrap = false;
for_multi_wrap = false;
current_time = GetCurrentTimestamp();
foreach (cell, dblist) {
avw_dbase* tmp = (avw_dbase*)lfirst(cell);
Dlelem* elem = NULL;
if (TransactionIdPrecedes(tmp->adw_frozenxid, xidForceLimit)) {
if (avdb == NULL || TransactionIdPrecedes(tmp->adw_frozenxid, avdb->adw_frozenxid))
avdb = tmp;
for_xid_wrap = true;
continue;
} else if (for_xid_wrap)
continue;
#ifndef ENABLE_MULTIPLE_NODES
else if (MultiXactIdPrecedes(tmp->adw_frozenmulti, multiForceLimit)) {
if (avdb == NULL || MultiXactIdPrecedes(tmp->adw_frozenmulti, avdb->adw_frozenmulti))
avdb = tmp;
for_multi_wrap = true;
continue;
} else if (for_multi_wrap)
continue;
#endif
tmp->adw_entry = pgstat_fetch_stat_dbentry(tmp->adw_datid);
* Skip a database with no pgstat entry; it means it hasn't seen any
* activity.
*/
if (NULL == tmp->adw_entry)
continue;
* Also, skip a database that appears on the database list as having
* been processed recently (less than autovacuum_naptime seconds ago).
* We do this so that we don't select a database which we just
* selected, but that pgstat hasn't gotten around to updating the last
* autovacuum time yet.
*/
skipit = false;
elem = t_thrd.autovacuum_cxt.DatabaseList ? DLGetTail(t_thrd.autovacuum_cxt.DatabaseList) : NULL;
while (elem != NULL) {
avl_dbase* dbp = (avl_dbase*)DLE_VAL(elem);
if (dbp->adl_datid == tmp->adw_datid) {
* Skip this database if its next_worker value falls between
* the current time and the current time plus naptime.
*/
if (!TimestampDifferenceExceeds(dbp->adl_next_worker, current_time, 0) &&
!TimestampDifferenceExceeds(
current_time, dbp->adl_next_worker, u_sess->attr.attr_storage.autovacuum_naptime * 1000))
skipit = true;
break;
}
elem = DLGetPred(elem);
}
if (skipit)
continue;
* Remember the db with oldest autovac time. (If we are here, both
* tmp->entry and db->entry must be non-null.)
*/
if (avdb == NULL || tmp->adw_entry->last_autovac_time < avdb->adw_entry->last_autovac_time)
avdb = tmp;
}
if (avdb != NULL) {
WorkerInfo worker = NULL;
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
* Get a worker entry from the freelist. We checked above, so there
* really should be a free slot -- complain very loudly if there
* isn't.
*/
worker = t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers;
if (worker == NULL)
ereport(FATAL, (errmsg("no free worker found")));
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers = (WorkerInfo)worker->wi_links.next;
worker->wi_dboid = avdb->adw_datid;
worker->wi_proc = NULL;
worker->wi_launchtime = GetCurrentTimestamp();
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker = worker;
LWLockRelease(AutovacuumLock);
SendPostmasterSignal(PMSIGNAL_START_AUTOVAC_WORKER);
retval = avdb->adw_datid;
} else if (skipit) {
* If we skipped all databases on the list, rebuild it, because it
* probably contains a dropped database.
*/
rebuild_database_list(InvalidOid);
}
(void)MemoryContextSwitchTo(oldcxt);
MemoryContextDelete(tmpcxt);
return retval;
}
* launch_worker
*
* Wrapper for starting a worker from the launcher. Besides actually starting
* it, update the database list to reflect the next time that another one will
* need to be started on the selected database. The actual database choice is
* left to do_start_worker.
*
* This routine is also expected to insert an entry into the database list if
* the selected database was previously absent from the list.
*/
static void launch_worker(TimestampTz now)
{
Oid dbid;
Dlelem* elem = NULL;
dbid = do_start_worker();
if (OidIsValid(dbid)) {
* Walk the database list and update the corresponding entry. If the
* database is not on the list, we'll recreate the list.
*/
elem = (t_thrd.autovacuum_cxt.DatabaseList == NULL) ? NULL : DLGetHead(t_thrd.autovacuum_cxt.DatabaseList);
while (elem != NULL) {
avl_dbase* avdb = (avl_dbase*)DLE_VAL(elem);
if (avdb->adl_datid == dbid) {
* add autovacuum_naptime seconds to the current time, and use
* that as the new "next_worker" field for this database.
*/
avdb->adl_next_worker =
TimestampTzPlusMilliseconds(now, u_sess->attr.attr_storage.autovacuum_naptime * 1000);
DLMoveToFront(elem);
break;
}
elem = DLGetSucc(elem);
}
* If the database was not present in the database list, we rebuild
* the list. It's possible that the database does not get into the
* list anyway, for example if it's a database that doesn't have a
* pgstat entry, but this is not a problem because we don't want to
* schedule workers regularly into those in any case.
*/
if (elem == NULL)
rebuild_database_list(dbid);
}
}
* Called from postmaster to signal a failure to fork a process to become
* worker. The postmaster should kill(SIGUSR2) the launcher shortly
* after calling this function.
*/
void AutoVacWorkerFailed(void)
{
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_signal[AutoVacForkFailed] = true;
}
static void avl_sighup_handler(SIGNAL_ARGS)
{
int save_errno = errno;
t_thrd.autovacuum_cxt.got_SIGHUP = true;
if (t_thrd.proc)
SetLatch(&t_thrd.proc->procLatch);
errno = save_errno;
}
static void avl_sigusr2_handler(SIGNAL_ARGS)
{
int save_errno = errno;
t_thrd.autovacuum_cxt.got_SIGUSR2 = true;
if (t_thrd.proc)
SetLatch(&t_thrd.proc->procLatch);
errno = save_errno;
}
static void avl_sigterm_handler(SIGNAL_ARGS)
{
int save_errno = errno;
t_thrd.autovacuum_cxt.got_SIGTERM = true;
if (t_thrd.proc)
SetLatch(&t_thrd.proc->procLatch);
errno = save_errno;
}
* AUTOVACUUM WORKER CODE
********************************************************************/
* AutoVacWorkerMain
*/
NON_EXEC_STATIC void AutoVacWorkerMain()
{
sigjmp_buf local_sigjmp_buf;
Oid dbid;
char user[NAMEDATALEN];
IsUnderPostmaster = true;
t_thrd.role = AUTOVACUUM_WORKER;
t_thrd.proc_cxt.MyProcPid = gs_thread_self();
t_thrd.proc_cxt.MyStartTime = time(NULL);
t_thrd.proc_cxt.MyProgName = "AutoVacWorker";
init_ps_display("autovacuum worker process", "", "", "");
SetProcessingMode(InitProcessing);
* Set up signal handlers. We operate on databases much like a regular
* backend, so we use the same signal handling. See equivalent code in
* tcop/postgres.c.
*
* Currently, we don't pay attention to postgresql.conf changes that
* happen during a single daemon iteration, so we can ignore SIGHUP.
*/
* SIGINT is used to signal canceling the current table's vacuum; SIGTERM
* means abort and exit cleanly, and SIGQUIT means abandon ship.
*/
(void)gspqsignal(SIGINT, StatementCancelHandler);
(void)gspqsignal(SIGTERM, die);
(void)gspqsignal(SIGQUIT, quickdie);
(void)gspqsignal(SIGALRM, handle_sig_alarm);
(void)gspqsignal(SIGPIPE, SIG_IGN);
(void)gspqsignal(SIGUSR1, procsignal_sigusr1_handler);
(void)gspqsignal(SIGUSR2, SIG_IGN);
(void)gspqsignal(SIGFPE, FloatExceptionHandler);
(void)gspqsignal(SIGCHLD, SIG_DFL);
(void)gspqsignal(SIGHUP, SIG_IGN);
(void)gspqsignal(SIGURG, print_stack);
BaseInit();
* Create a per-backend PGPROC struct in shared memory, except in the
* EXEC_BACKEND case where this was done in SubPostmasterMain. We must do
* this before we can use LWLocks (and in the EXEC_BACKEND case we already
* had to do some stuff with LWLocks).
*/
#ifndef EXEC_BACKEND
InitProcess();
#endif
on_proc_exit(autoVacQuitAndClean, 0);
* If an exception is encountered, processing resumes here.
*
* See notes in postgres.c about the design of this coding.
*/
int curTryCounter;
int* oldTryCounter = NULL;
if (sigsetjmp(local_sigjmp_buf, 1) != 0) {
gstrace_tryblock_exit(true, oldTryCounter);
HOLD_INTERRUPTS();
EmitErrorReport();
AtEOXact_SysDBCache(false);
* We can now go away. Note that because we called InitProcess, a
* callback was registered to do ProcKill, which will clean up
* necessary state.
*/
proc_exit(0);
}
oldTryCounter = gstrace_tryblock_entry(&curTryCounter);
t_thrd.log_cxt.PG_exception_stack = &local_sigjmp_buf;
gs_signal_setmask(&t_thrd.libpq_cxt.UnBlockSig, NULL);
(void)gs_signal_unblock_sigusr2();
* Force zero_damaged_pages OFF in the autovac process, even if it is set
* in postgresql.conf. We don't really want such a dangerous option being
* applied non-interactively.
*/
SetConfigOption("zero_damaged_pages", "false", PGC_SUSET, PGC_S_OVERRIDE);
* Force statement_timeout to zero to avoid a timeout setting from
* preventing regular maintenance from being executed.
*/
SetConfigOption("statement_timeout", "0", PGC_SUSET, PGC_S_OVERRIDE);
* Force default_transaction_isolation to READ COMMITTED. We don't want
* to pay the overhead of serializable mode, nor add any risk of causing
* deadlocks or delaying other transactions.
*/
SetConfigOption("default_transaction_isolation", "read committed", PGC_SUSET, PGC_S_OVERRIDE);
* Get the info about the database we're going to work on.
*/
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
* beware of startingWorker being INVALID; this should normally not
* happen, but if a worker fails after forking and before this, the
* launcher might have decided to remove it from the queue and start
* again.
*/
if (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker != NULL) {
t_thrd.autovacuum_cxt.MyWorkerInfo = t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker;
dbid = t_thrd.autovacuum_cxt.MyWorkerInfo->wi_dboid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_proc = t_thrd.proc;
SHMQueueInsertBefore(
&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers, &t_thrd.autovacuum_cxt.MyWorkerInfo->wi_links);
* remove from the "starting" pointer, so that the launcher can start
* a new worker if required
*/
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker = NULL;
LWLockRelease(AutovacuumLock);
on_shmem_exit(FreeWorkerInfo, 0);
on_shmem_exit(PGXCNodeCleanAndRelease, 0);
if (t_thrd.autovacuum_cxt.AutoVacuumShmem->av_launcherpid != 0)
gs_signal_send(t_thrd.autovacuum_cxt.AutoVacuumShmem->av_launcherpid, SIGUSR2);
} else {
ereport(WARNING, (errmsg("autovacuum worker started without a worker entry")));
dbid = InvalidOid;
LWLockRelease(AutovacuumLock);
}
if (OidIsValid(dbid)) {
char dbname[NAMEDATALEN];
MemoryContext oldcontext = NULL;
* Report autovac startup to the stats collector. We deliberately do
* this before Init openGauss, so that the last_autovac_time will get
* updated even if the connection attempt fails. This is to prevent
* autovac from getting "stuck" repeatedly selecting an unopenable
* database, rather than making any progress on stuff it can connect
* to.
*/
pgstat_report_autovac(dbid);
AUTOVAC_LOG(LOG, "report autovac startup on database %u to stats collector", dbid);
* Connect to the selected database
*
* Note: if we have selected a just-deleted database (due to using
* stale stats info), we'll fail and exit here.
*/
t_thrd.proc_cxt.PostInit->SetDatabaseAndUser(NULL, dbid, NULL);
t_thrd.proc_cxt.PostInit->InitAutoVacWorker();
t_thrd.proc_cxt.PostInit->GetDatabaseName(dbname);
#ifndef ENABLE_MULTIPLE_NODES
AutoDopControl dopControl;
dopControl.CloseSmp();
#endif
SetProcessingMode(NormalProcessing);
pgstat_report_appname("AutoVacWorker");
pgstat_report_activity(STATE_IDLE, NULL);
set_ps_display(dbname, false);
ereport(GetVacuumLogLevel(), (errmsg("start autovacuum on database \"%s\"", dbname)));
if (u_sess->attr.attr_security.PostAuthDelay)
pg_usleep(u_sess->attr.attr_security.PostAuthDelay * 1000000L);
* Create the memory context we will use in the main loop.
*
* t_thrd.mem_cxt.msg_mem_cxt is reset once per iteration of the main loop, ie, upon
* completion of processing of each command message from the client.
*/
t_thrd.mem_cxt.msg_mem_cxt = AllocSetContextCreate(t_thrd.top_mem_cxt, "MessageContext",
ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MAXSIZE);
t_thrd.mem_cxt.mask_password_mem_cxt = AllocSetContextCreate(t_thrd.top_mem_cxt, "MaskPasswordCtx",
ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MAXSIZE);
* Create a resource owner to keep track of our resources (currently only
* buffer pins).
*/
t_thrd.utils_cxt.CurrentResourceOwner = ResourceOwnerCreate(NULL, "AutoVacuumWorker",
THREAD_GET_MEM_CXT_GROUP(MEMORY_CONTEXT_STORAGE));
#if (!defined(ENABLE_MULTIPLE_NODES)) && (!defined(ENABLE_PRIVATEGAUSS))
LoadSqlPlugin();
#endif
oldcontext = MemoryContextSwitchTo(SESS_GET_MEM_CXT_GROUP(MEMORY_CONTEXT_STORAGE));
if (u_sess->proc_cxt.MyProcPort->database_name)
pfree_ext(u_sess->proc_cxt.MyProcPort->database_name);
if (u_sess->proc_cxt.MyProcPort->user_name)
pfree_ext(u_sess->proc_cxt.MyProcPort->user_name);
u_sess->proc_cxt.MyProcPort->database_name = pstrdup(dbname);
u_sess->proc_cxt.MyProcPort->user_name = (char*)GetSuperUserName((char*)user);
(void)MemoryContextSwitchTo(oldcontext);
exec_init_poolhandles();
t_thrd.autovacuum_cxt.recentXid = ReadNewTransactionId();
t_thrd.autovacuum_cxt.recentMulti = ReadNextMultiXactId();
do_autovacuum();
}
* The launcher will be notified of my death in ProcKill, *if* we managed
* to get a worker slot at all
*/
proc_exit(0);
}
* Return a WorkerInfo to the free list
*/
static void FreeWorkerInfo(int code, Datum arg)
{
if (t_thrd.autovacuum_cxt.MyWorkerInfo != NULL) {
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
if (t_thrd.autovacuum_cxt.MyWorkerInfo == NULL) {
LWLockRelease(AutovacuumLock);
return;
}
* Wake the launcher up so that he can launch a new worker immediately
* if required. We only save the launcher's PID in local memory here;
* the actual signal will be sent when the PGPROC is recycled. Note
* that we always do this, so that the launcher can rebalance the cost
* limit setting of the remaining workers.
*
* We somewhat ignore the risk that the launcher changes its PID
* between us reading it and the actual kill; we expect ProcKill to be
* called shortly after us, and we assume that PIDs are not reused too
* quickly after a process exits.
*/
t_thrd.autovacuum_cxt.AutovacuumLauncherPid = t_thrd.autovacuum_cxt.AutoVacuumShmem->av_launcherpid;
SHMQueueDelete(&t_thrd.autovacuum_cxt.MyWorkerInfo->wi_links);
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_links.next =
(SHM_QUEUE*)t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_dboid = InvalidOid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_tableoid = InvalidOid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_parentoid = InvalidOid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_sharedrel = false;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_proc = NULL;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_launchtime = 0;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_delay = 0;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_limit = 0;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_limit_base = 0;
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers = t_thrd.autovacuum_cxt.MyWorkerInfo;
t_thrd.autovacuum_cxt.MyWorkerInfo = NULL;
* now that we're inactive, cause a rebalancing of the surviving
* workers
*/
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_signal[AutoVacRebalance] = true;
LWLockRelease(AutovacuumLock);
}
}
* Update the cost-based delay parameters, so that multiple workers consume
* each a fraction of the total available I/O.
*/
void AutoVacuumUpdateDelay(void)
{
if (t_thrd.autovacuum_cxt.MyWorkerInfo) {
u_sess->attr.attr_storage.VacuumCostDelay = t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_delay;
u_sess->attr.attr_storage.VacuumCostLimit = t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_limit;
}
}
* autovac_balance_cost
* Recalculate the cost limit setting for each active worker.
*
* Caller must hold the AutovacuumLock in exclusive mode.
*/
static void autovac_balance_cost(void)
{
* The idea here is that we ration out I/O equally. The amount of I/O
* that a worker can consume is determined by cost_limit/cost_delay, so we
* try to equalize those ratios rather than the raw limit settings.
*
* note: in cost_limit, zero also means use value from elsewhere, because
* zero is not a valid value.
*/
int vac_cost_limit =
(u_sess->attr.attr_storage.autovacuum_vac_cost_limit > 0 ? u_sess->attr.attr_storage.autovacuum_vac_cost_limit
: u_sess->attr.attr_storage.VacuumCostLimit);
int vac_cost_delay =
(u_sess->attr.attr_storage.autovacuum_vac_cost_delay >= 0 ? u_sess->attr.attr_storage.autovacuum_vac_cost_delay
: u_sess->attr.attr_storage.VacuumCostDelay);
double cost_total;
double cost_avail;
WorkerInfo worker = NULL;
if (vac_cost_limit <= 0 || vac_cost_delay <= 0) {
return;
}
cost_total = 0.0;
worker = (WorkerInfo)SHMQueueNext(&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers,
&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers, offsetof(WorkerInfoData, wi_links));
while (worker != NULL) {
if (worker->wi_proc != NULL && worker->wi_cost_limit_base > 0 && worker->wi_cost_delay > 0)
cost_total += (double)worker->wi_cost_limit_base / worker->wi_cost_delay;
worker = (WorkerInfo)SHMQueueNext(&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers, &worker->wi_links,
offsetof(WorkerInfoData, wi_links));
}
if (cost_total <= 0) {
return;
}
* Adjust cost limit of each active worker to balance the total of cost
* limit to autovacuum_vacuum_cost_limit.
*/
cost_avail = (double)vac_cost_limit / vac_cost_delay;
worker = (WorkerInfo)SHMQueueNext(&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers,
&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers, offsetof(WorkerInfoData, wi_links));
while (worker != NULL) {
if (worker->wi_proc != NULL && worker->wi_cost_limit_base > 0 && worker->wi_cost_delay > 0) {
int limit = (int)(cost_avail * worker->wi_cost_limit_base / cost_total);
* We put a lower bound of 1 on the cost_limit, to avoid division-
* by-zero in the vacuum code. Also, in case of roundoff trouble
* in these calculations, let's be sure we don't ever set
* cost_limit to more than the base value.
*/
worker->wi_cost_limit = Max(Min(limit, worker->wi_cost_limit_base), 1);
AUTOVAC_LOG(LOG, "autovac_balance_cost(pid=%lu db=%u, rel=%u, parent=%u, cost_limit=%d, "
"cost_limit_base=%d, cost_delay=%d)",
worker->wi_proc->pid, worker->wi_dboid, worker->wi_tableoid, worker->wi_parentoid,
worker->wi_cost_limit, worker->wi_cost_limit_base, worker->wi_cost_delay);
}
worker = (WorkerInfo)SHMQueueNext(&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers,
&worker->wi_links,
offsetof(WorkerInfoData, wi_links));
}
}
* get_database_list
* Return a list of all databases found in pg_database.
*
* The list and associated data is allocated in the caller's memory context,
* which is in charge of ensuring that it's properly cleaned up afterwards.
*
* Note: this is the only function in which the autovacuum launcher uses a
* transaction. Although we aren't attached to any particular database and
* therefore can't access most catalogs, we do have enough infrastructure
* to do a seqscan on pg_database.
*/
static List* get_database_list(void)
{
List* dblist = NIL;
Relation rel;
TableScanDesc scan;
HeapTuple tup;
MemoryContext resultcxt;
resultcxt = CurrentMemoryContext;
* Start a transaction so we can access pg_database, and get a snapshot.
* We don't have a use for the snapshot itself, but we're interested in
* the secondary effect that it sets RecentGlobalXmin. (This is critical
* for anything that reads heap pages, because HOT may decide to prune
* them even if the process doesn't attempt to modify any tuples.)
*/
StartTransactionCommand();
(void)GetTransactionSnapshot();
rel = heap_open(DatabaseRelationId, AccessShareLock);
scan = tableam_scan_begin(rel, SnapshotNow, 0, NULL);
while (HeapTupleIsValid(tup = (HeapTuple) tableam_scan_getnexttuple(scan, ForwardScanDirection))) {
Form_pg_database pgdatabase = (Form_pg_database)GETSTRUCT(tup);
avw_dbase* avdb = NULL;
MemoryContext oldcxt;
* Allocate our results in the caller's context, not the
* transaction's. We do this inside the loop, and restore the original
* context at the end, so that leaky things like heap_getnext() are
* not called in a potentially long-lived context.
*/
oldcxt = MemoryContextSwitchTo(resultcxt);
avdb = (avw_dbase*)palloc(sizeof(avw_dbase));
avdb->adw_datid = HeapTupleGetOid(tup);
avdb->adw_name = pstrdup(NameStr(pgdatabase->datname));
bool isNull = false;
TransactionId datfrozenxid;
Datum xid64datum = heap_getattr(tup, Anum_pg_database_datfrozenxid64, RelationGetDescr(rel), &isNull);
if (isNull) {
datfrozenxid = pgdatabase->datfrozenxid;
if (TransactionIdPrecedes(t_thrd.xact_cxt.ShmemVariableCache->nextXid, datfrozenxid))
datfrozenxid = FirstNormalTransactionId;
} else
datfrozenxid = DatumGetTransactionId(xid64datum);
#ifndef ENABLE_MULTIPLE_NODES
Datum mxidDatum = heap_getattr(tup, Anum_pg_database_datminmxid, RelationGetDescr(rel), &isNull);
MultiXactId datminmxid = isNull ? FirstMultiXactId : DatumGetTransactionId(mxidDatum);
avdb->adw_frozenmulti = datminmxid;
#endif
avdb->adw_frozenxid = datfrozenxid;
avdb->adw_entry = NULL;
dblist = lappend(dblist, avdb);
(void)MemoryContextSwitchTo(oldcxt);
}
tableam_scan_end(scan);
heap_close(rel, AccessShareLock);
CommitTransactionCommand();
return dblist;
}
* 1. just support global statistic since local statistic is gradually abandoned
* 2. just support traditional-sample analyze since percent-sample analyze is too inefficient
*/
#define DO_ANALYZE \
((AUTOVACUUM_DO_ANALYZE_VACUUM == u_sess->attr.attr_storage.autovacuum_mode || \
AUTOVACUUM_DO_ANALYZE == u_sess->attr.attr_storage.autovacuum_mode) && \
u_sess->attr.attr_sql.enable_global_stats && 0 < default_statistics_target)
#define DO_VACUUM \
(AUTOVACUUM_DO_ANALYZE_VACUUM == u_sess->attr.attr_storage.autovacuum_mode || \
AUTOVACUUM_DO_VACUUM == u_sess->attr.attr_storage.autovacuum_mode)
* check if the relation can do auto-analyze or auto-vacuum
*/
void relation_support_autoavac(HeapTuple tuple, bool* enable_analyze, bool* enable_vacuum, bool* is_internal_relation)
{
bytea* relopts = NULL;
Form_pg_class classForm = (Form_pg_class)GETSTRUCT(tuple);
Assert(PointerIsValid(enable_analyze));
Assert(PointerIsValid(enable_vacuum));
*enable_analyze = false;
*enable_vacuum = false;
*is_internal_relation = false;
if (IS_PGXC_COORDINATOR && !u_sess->attr.attr_storage.autovacuum_start_daemon)
return;
* 1. never analyze internal table since we never select them directly
* 2. just analyze row/colume orientation table since analyze hdfs table is too inefficient
*/
relopts = extractRelOptions(tuple, GetDefaultPgClassDesc(), InvalidOid);
if (StdRelOptGetInternalMask(relopts)) {
*is_internal_relation = true;
if (StdRelOptIsRowStore(relopts) && (ISMATMAP(classForm->relname.data) || ISMLOG(classForm->relname.data))) {
*enable_analyze = true;
*enable_vacuum = true;
}
} else if (StdRelOptIsColStore(relopts)) {
*enable_analyze = true;
} else if (StdRelOptIsRowStore(relopts)) {
*enable_analyze = true;
*enable_vacuum = true;
}
if (StatisticRelationId == HeapTupleGetOid(tuple) || RELKIND_TOASTVALUE == classForm->relkind || !DO_ANALYZE) {
*enable_analyze = false;
}
if (!DO_VACUUM) {
*enable_vacuum = false;
}
if (relopts != NULL)
pfree_ext(relopts);
* 1. data in temp/unlogged is short-lived, so do nothing for temp/unlogged table
* 2. foreign table dose not have stat info, so just support ordinary table
*/
if (RELPERSISTENCE_PERMANENT != classForm->relpersistence || RELKIND_RELATION != classForm->relkind) {
*enable_analyze = false;
*is_internal_relation = false;
}
if (RELKIND_RELATION != classForm->relkind
#ifndef ENABLE_MULTIPLE_NODES
&& RELKIND_TOASTVALUE != classForm->relkind
#endif
) {
*enable_vacuum = false;
}
}
bool allow_autoanalyze(HeapTuple tuple)
{
bool enable_analyze = false;
bool enable_vacuum = false;
bool is_internal_relation = false;
relation_support_autoavac(tuple, &enable_analyze, &enable_vacuum, &is_internal_relation);
return enable_analyze;
}
static void AddApplicationNameToPoolerParams()
{
if (IS_PGXC_COORDINATOR && !IsConnFromCoord() && IsAutoVacuumWorkerProcess()) {
StringInfoData str;
initStringInfo(&str);
appendStringInfo(&str, "SET application_name = '%s';", AUTO_VACUUM_WORKER);
(void)register_pooler_session_param("application_name", str.data, POOL_CMD_GLOBAL_SET);
pfree_ext(str.data);
}
}
static void DeleteApplicationNameFromPoolerParams()
{
if (IS_PGXC_COORDINATOR && !IsConnFromCoord() && IsAutoVacuumWorkerProcess()) {
delete_pooler_session_params("application_name");
}
}
static void fetch_global_autovac_info()
{
PgStat_StatTabKey tablekey;
bool connected = false;
StringInfoData buf;
* Dose not fetch global stat info from all datanodes if
* 1. autovacuum = off
* 2. autovacuum = on and u_sess->attr.attr_storage.autovacuum_mode = none
*/
if (!u_sess->attr.attr_storage.autovacuum_start_daemon ||
AUTOVACUUM_DO_NONE == u_sess->attr.attr_storage.autovacuum_mode)
return;
initStringInfo(&buf);
if (DO_VACUUM) {
appendStringInfo(&buf,
"with f as (select nspname, relname, partname, pg_catalog.sum(n_dead_tuples) as n_dead_tuples, "
"pg_catalog.sum(n_live_tuples) as n_live_tuples, pg_catalog.sum(changes_since_analyze) "
"as changes_since_analyze, pg_catalog.count(1) as count from "
"pg_catalog.pg_total_autovac_tuples(%s) group by nspname, relname, partname), "
"t as(SELECT c.oid as relid,n.nspname AS nspname, c.relname AS relname, "
"case when p.parttype = 'r' then null else p.oid end as partid, "
"case when p.parttype = 'r' then null else p.relname end as partname, x.pclocatortype as locatortype "
"FROM pg_class c INNER JOIN pg_namespace n ON n.oid = c.relnamespace INNER JOIN pgxc_class x "
"on x.pcrelid = c.oid LEFT JOIN pg_partition p on c.oid = p.parentid "
"WHERE c.relkind = 'r' and n.nspname not in ('pg_toast','cstore'))"
"select t.relid, t.partid, (case when locatortype = 'R' then (f.n_dead_tuples/f.count) else "
"f.n_dead_tuples end)::bigint as n_dead_tuples, "
"(case when locatortype = 'R' then (f.n_live_tuples/f.count) else f.n_live_tuples end)::bigint as "
"n_live_tuples, (case when locatortype = 'R' then "
"(f.changes_since_analyze/f.count) else f.changes_since_analyze end)::bigint as changes_since_analyze "
"from t inner join f on (t.nspname = f.nspname and t.relname = f.relname "
"and (t.partname = f.partname or (t.partname is null and f.partname is null))) ", "false");
} else {
appendStringInfo(&buf, "with f as (select nspname, relname, pg_catalog.sum(n_dead_tuples) as n_dead_tuples, "
"pg_catalog.sum(changes_since_analyze) as changes_since_analyze, pg_catalog.count(1) as count "
"from pg_catalog.pg_total_autovac_tuples(%s) group by nspname, relname), t as(SELECT c.oid as relid, "
"n.nspname AS nspname, c.relname AS relname, x.pclocatortype as locatortype FROM pg_class c "
"INNER JOIN pg_namespace n ON n.oid = c.relnamespace INNER JOIN pgxc_class x on x.pcrelid = c.oid "
"WHERE c.relkind = 'r' and c.relpersistence = 'p' and n.nspname not in ('pg_toast','cstore'))"
"select t.relid, NULL AS partid, "
"(case when locatortype = 'R' then (f.n_dead_tuples/f.count) else f.n_dead_tuples end)::bigint as "
"n_dead_tuples, 0 AS n_live_tuples, "
"(case when locatortype = 'R' then (f.changes_since_analyze/f.count) else f.changes_since_analyze "
"end)::bigint as changes_since_analyze "
"from t inner join f on (t.nspname = f.nspname and t.relname = f.relname) ", "true");
}
AUTOVAC_LOG(DEBUG2, "FETCH GLOABLE AUTOVAC INFO STRING: %s", buf.data);
PushActiveSnapshot(GetTransactionSnapshot());
AddApplicationNameToPoolerParams();
PG_TRY();
{
DEBUG_MOD_START_TIMER(MOD_AUTOVAC);
SPI_STACK_LOG("connect", NULL, NULL);
if (SPI_OK_CONNECT != SPI_connect()) {
ereport(ERROR, (errcode(ERRCODE_OPERATE_FAILED), errmsg("Unable to connect to execute internal query.")));
}
connected = true;
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: build SPI connect");
if (SPI_OK_SELECT != SPI_execute(buf.data, true, 0)) {
ereport(ERROR, (errcode(ERRCODE_DATA_EXCEPTION), errmsg("fail to execute query")));
}
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: execte SQL to fetch global autovac info");
pfree_ext(buf.data);
for (uint32 i = 0; i < SPI_processed; i++) {
Oid relid;
Oid partid;
int64 n_dead_tuples = 0;
int64 n_live_tuples = 0;
int64 changes_since_analyze = 0;
bool isnull = false;
bool partid_isnull = true;
avw_info* entry = NULL;
relid = DatumGetObjectId(SPI_getbinval(SPI_tuptable->vals[i], SPI_tuptable->tupdesc, 1, &isnull));
partid = DatumGetObjectId(SPI_getbinval(SPI_tuptable->vals[i], SPI_tuptable->tupdesc, 2, &partid_isnull));
n_dead_tuples = DatumGetInt64(SPI_getbinval(SPI_tuptable->vals[i], SPI_tuptable->tupdesc, 3, &isnull));
n_live_tuples = DatumGetInt64(SPI_getbinval(SPI_tuptable->vals[i], SPI_tuptable->tupdesc, 4, &isnull));
changes_since_analyze =
DatumGetInt64(SPI_getbinval(SPI_tuptable->vals[i], SPI_tuptable->tupdesc, 5, &isnull));
if (partid_isnull) {
tablekey.tableid = relid;
tablekey.statFlag = InvalidOid;
} else {
tablekey.tableid = partid;
tablekey.statFlag = relid;
}
entry =
(avw_info*)hash_search(t_thrd.autovacuum_cxt.pgStatAutoVacInfo, (void*)(&tablekey), HASH_ENTER, NULL);
entry->n_dead_tuples = n_dead_tuples;
entry->n_live_tuples = n_live_tuples;
entry->changes_since_analyze = changes_since_analyze;
}
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: process %u SPI tuples", SPI_processed);
connected = false;
SPI_STACK_LOG("finish", NULL, NULL);
if (SPI_OK_FINISH != SPI_finish()) {
ereport(ERROR, (errcode(ERRCODE_OPERATE_FAILED), errmsg("SPI_finish failed")));
}
PopActiveSnapshot();
u_sess->debug_query_id = 0;
}
PG_CATCH();
{
DeleteApplicationNameFromPoolerParams();
if (connected) {
SPI_STACK_LOG("finish", NULL, NULL);
SPI_finish();
}
PopActiveSnapshot();
PG_RE_THROW();
u_sess->debug_query_id = 0;
}
PG_END_TRY();
DeleteApplicationNameFromPoolerParams();
}
static void ClearStatsHistoryTimeoutRecords()
{
ScanKeyData skey[1];
SysScanDesc sysscan;
HeapTuple tuple;
TimestampTz retentionTime;
TimestampTz currentTimestamp = GetCurrentTimestamp();
int retention = u_sess->attr.attr_sql.gms_stats_history_retention;
#ifdef HAVE_INT64_TIMESTAMP
retentionTime = retention * USECS_PER_DAY;
#else
retentionTime = retention * SECS_PER_DAY;
#endif
ScanKeyInit(&skey[0], Anum_pg_statistic_history_current_analyzetime, BTLessStrategyNumber, F_TIMESTAMP_LT,
TimestampTzGetDatum(currentTimestamp - retentionTime));
Relation rel = heap_open(StatisticHistoryRelationId, RowExclusiveLock);
sysscan = systable_beginscan(rel, StatisticHistoryCurrTimeRelidIndexId, true, SnapshotNow, 1, skey);
while (HeapTupleIsValid(tuple = systable_getnext(sysscan))) {
simple_heap_delete(rel, &tuple->t_self);
}
systable_endscan(sysscan);
heap_close(rel, RowExclusiveLock);
}
static bool check_table_is_locked(autovac_table* tab)
{
if (t_thrd.proc->workingVersionNum < STATISTIC_HISTORY_VERSION_NUMBER) {
return false;
}
Oid namespaceid = get_rel_namespace(tab->at_relid);
return CheckRelationLocked(namespaceid, tab->at_relid);
}
* Process a database table-by-table
*
* Note that CHECK_FOR_INTERRUPTS is supposed to be used in certain spots in
* order not to ignore shutdown commands for too long.
*/
static void do_autovacuum(void)
{
Relation classRel = NULL;
HeapTuple tuple = NULL;
TableScanDesc relScan = NULL;
Form_pg_database dbForm = NULL;
List* table_oids = NIL;
HASHCTL partitioned_tables_ctl;
HTAB* partitioned_tables_map = NULL;
HASHCTL table_relopt_ctl;
HTAB* table_relopt_map = NULL;
HASHCTL toast_table_ctl;
HTAB* toast_table_map = NULL;
ListCell* volatile cell = NULL;
PgStat_StatDBEntry* shared = NULL;
PgStat_StatDBEntry* dbentry = NULL;
BufferAccessStrategy bstrategy;
HASHCTL avinfo_ctl;
bool datallowconn = true;
bool local_autovacuum = true;
bool freeze_autovacuum = false;
bool is_internal_relation = false;
ScanKeyData key[1];
TableScanDesc partScan;
TableScanDesc subpartScan;
Relation partRel;
HeapTuple partTuple;
TupleDesc pg_class_desc;
vacuum_object* vacObj = NULL;
errno_t rc = EOK;
* StartTransactionCommand and CommitTransactionCommand will automatically
* switch to other contexts. We need this one to keep the list of
* relations to vacuum/analyze across transactions.
*/
t_thrd.autovacuum_cxt.AutovacMemCxt = AllocSetContextCreate(
t_thrd.top_mem_cxt, "AV worker", ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MAXSIZE);
(void)MemoryContextSwitchTo(t_thrd.autovacuum_cxt.AutovacMemCxt);
* may be NULL if we couldn't find an entry (only happens if we are
* forcing a vacuum for anti-wrap purposes).
*/
dbentry = pgstat_fetch_stat_dbentry(u_sess->proc_cxt.MyDatabaseId);
StartTransactionCommand();
* Clean up any dead statistics collector entries for this DB. We always
* want to do this exactly once per DB-processing cycle, even if we find
* nothing worth vacuuming in the database.
*/
DEBUG_MOD_START_TIMER(MOD_AUTOVAC);
pgstat_vacuum_stat();
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: Clean up dead statistics collector entries for current DB");
* Find the pg_database entry and select the default freeze ages. We use
* zero in template and nonconnectable databases, else the system-wide
* default.
*/
tuple = SearchSysCache1(DATABASEOID, ObjectIdGetDatum(u_sess->proc_cxt.MyDatabaseId));
if (!HeapTupleIsValid(tuple))
ereport(ERROR, (errcode(ERRCODE_CACHE_LOOKUP_FAILED),
errmsg("cache lookup failed for database %u", u_sess->proc_cxt.MyDatabaseId)));
dbForm = (Form_pg_database)GETSTRUCT(tuple);
datallowconn = dbForm->datallowconn;
if (dbForm->datistemplate || !dbForm->datallowconn) {
t_thrd.autovacuum_cxt.default_freeze_min_age = 0;
t_thrd.autovacuum_cxt.default_freeze_table_age = 0;
} else {
t_thrd.autovacuum_cxt.default_freeze_min_age = u_sess->attr.attr_storage.vacuum_freeze_min_age;
t_thrd.autovacuum_cxt.default_freeze_table_age = u_sess->attr.attr_storage.vacuum_freeze_table_age;
}
ReleaseSysCache(tuple);
#ifdef PGXC
if (u_sess->attr.attr_common.upgrade_mode == 1) {
CommitTransactionCommand();
return;
}
if (false == datallowconn || IS_SINGLE_NODE) {
local_autovacuum = true;
freeze_autovacuum = false;
} else if (IS_PGXC_COORDINATOR && PgxcIsCentralCoordinator(g_instance.attr.attr_common.PGXCNodeName)) {
local_autovacuum = false;
freeze_autovacuum = false;
} else {
local_autovacuum = true;
freeze_autovacuum = true;
}
#endif
(void)MemoryContextSwitchTo(t_thrd.autovacuum_cxt.AutovacMemCxt);
t_thrd.autovacuum_cxt.pgStatAutoVacInfo = NULL;
if (IS_PGXC_COORDINATOR && false == local_autovacuum) {
rc = memset_s(&avinfo_ctl, sizeof(avinfo_ctl), 0, sizeof(avinfo_ctl));
securec_check(rc, "", "");
avinfo_ctl.keysize = sizeof(PgStat_StatTabKey);
avinfo_ctl.entrysize = sizeof(avw_info);
avinfo_ctl.hcxt = t_thrd.autovacuum_cxt.AutovacMemCxt;
avinfo_ctl.hash = tag_hash;
t_thrd.autovacuum_cxt.pgStatAutoVacInfo = hash_create(
"autovac information of user-define table", 512, &avinfo_ctl, HASH_ELEM | HASH_FUNCTION | HASH_CONTEXT);
fetch_global_autovac_info();
}
shared = pgstat_fetch_stat_dbentry(InvalidOid);
classRel = heap_open(RelationRelationId, AccessShareLock);
pg_class_desc = CreateTupleDescCopy(RelationGetDescr(classRel));
rc = memset_s(&partitioned_tables_ctl, sizeof(partitioned_tables_ctl), 0, sizeof(partitioned_tables_ctl));
securec_check(rc, "", "");
partitioned_tables_ctl.keysize = sizeof(Oid);
partitioned_tables_ctl.entrysize = sizeof(at_partitioned_table);
partitioned_tables_ctl.hash = oid_hash;
partitioned_tables_map =
hash_create("partitioned relid to autovac info map", 10, &partitioned_tables_ctl, HASH_ELEM | HASH_FUNCTION);
rc = memset_s(&toast_table_ctl, sizeof(toast_table_ctl), 0, sizeof(toast_table_ctl));
securec_check(rc, "", "");
toast_table_ctl.keysize = sizeof(Oid);
toast_table_ctl.entrysize = sizeof(av_toastid_mainid);
toast_table_ctl.hash = oid_hash;
toast_table_map = hash_create("TOAST to main relid map", 100, &toast_table_ctl, HASH_ELEM | HASH_FUNCTION);
rc = memset_s(&table_relopt_ctl, sizeof(table_relopt_ctl), 0, sizeof(table_relopt_ctl));
securec_check(rc, "", "");
table_relopt_ctl.keysize = sizeof(Oid);
table_relopt_ctl.entrysize = sizeof(av_relation);
table_relopt_ctl.hash = oid_hash;
table_relopt_map = hash_create("main relid to rel options map", 100, &table_relopt_ctl, HASH_ELEM | HASH_FUNCTION);
* Scan pg_class to determine which tables to vacuum.
*
* relations and materialized views, and on the second one we collect
* TOAST tables. The reason for doing the second pass is that during it we
* want to use the main relation's pg_class.reloptions entry if the TOAST
* table does not have any, and we cannot obtain it unless we know
* beforehand what's the main table OID.
*
* We need to check TOAST tables separately because in cases with short,
* wide tables there might be proportionally much more activity in the
* TOAST table than in its parent.
*/
relScan = tableam_scan_begin(classRel, SnapshotNow, 0, NULL);
* On the first pass, we collect main tables to vacuum, and also the main
* table relid to TOAST relid mapping.
*/
while ((tuple = (HeapTuple) tableam_scan_getnexttuple(relScan, ForwardScanDirection)) != NULL) {
Form_pg_class classForm = (Form_pg_class)GETSTRUCT(tuple);
PgStat_StatTabEntry* tabentry = NULL;
AutoVacOpts* relopts = NULL;
Oid relid = HeapTupleGetOid(tuple);
bool dovacuum = false;
bool doanalyze = false;
bool need_freeze = false;
bool enable_analyze = false;
bool enable_vacuum = false;
if (classForm->relkind != RELKIND_RELATION &&
classForm->relkind != RELKIND_MATVIEW)
continue;
if (classForm->relpersistence == RELPERSISTENCE_TEMP ||
classForm->relpersistence == RELPERSISTENCE_GLOBAL_TEMP) {
continue;
}
* for Ustore partitioned tables
*/
bytea *rawRelopts = extractRelOptions(tuple, pg_class_desc, InvalidOid);
if (rawRelopts != NULL && RelationIsTableAccessMethodUStoreType(rawRelopts) &&
isPartitionedRelation(classForm)) {
vacObj = (vacuum_object*)palloc(sizeof(vacuum_object));
vacObj->tab_oid = relid;
vacObj->parent_oid = InvalidOid;
vacObj->dovacuum = true;
vacObj->dovacuum_toast = false;
vacObj->doanalyze = false;
vacObj->need_freeze = false;
vacObj->is_internal_relation = false;
vacObj->flags = VACFLG_MAIN_PARTITION;
table_oids = lappend(table_oids, vacObj);
continue;
}
relopts = extract_autovac_opts(tuple, pg_class_desc);
tabentry = get_pgstat_tabentry_relid(relid, classForm->relisshared, InvalidOid, shared, dbentry);
relation_support_autoavac(tuple, &enable_analyze, &enable_vacuum, &is_internal_relation);
relation_needs_vacanalyze(relid, relopts, rawRelopts, classForm, tuple, tabentry, enable_analyze, enable_vacuum,
false, &dovacuum, &doanalyze, &need_freeze);
if (freeze_autovacuum) {
dovacuum = need_freeze;
doanalyze = false;
}
if (dovacuum || doanalyze) {
vacObj = (vacuum_object*)palloc(sizeof(vacuum_object));
vacObj->tab_oid = relid;
vacObj->parent_oid = InvalidOid;
vacObj->dovacuum = dovacuum;
vacObj->dovacuum_toast = false;
vacObj->doanalyze = doanalyze;
vacObj->need_freeze = need_freeze;
vacObj->is_internal_relation = is_internal_relation;
vacObj->gpi_vacuumed = false;
vacObj->flags = (isPartitionedRelation(classForm) ? VACFLG_MAIN_PARTITION : VACFLG_SIMPLE_HEAP);
table_oids = lappend(table_oids, vacObj);
}
* record partitioned table's autovac stat
* 1. if we have to vaccum partitioned table since need freeze old tuple,
* we just skip vacuum its partition
* 2. to avoid recompute allowvacuum falg
*/
if (isPartitionedRelation(classForm)) {
bool found = false;
at_partitioned_table* ap_entry = NULL;
ap_entry = (at_partitioned_table*)hash_search(partitioned_tables_map, &relid, HASH_ENTER, &found);
if (!found) {
ap_entry->at_allowvacuum = enable_vacuum;
ap_entry->at_doanalyze = doanalyze;
ap_entry->at_dovacuum = dovacuum;
ap_entry->at_needfreeze = need_freeze;
ap_entry->at_gpivacuumed = false;
}
}
* Remember the association for the third pass. Note: we must do
* this even if the table is going to be vacuumed, because we
* don't automatically vacuum toast tables along the parent table.
*
* AutoVacOpts are recorded in partitioned table's reloptions, so we read
* and save AutoVacOpts so we can use them when we deal with parttition
*/
if (OidIsValid(classForm->reltoastrelid) || isPartitionedRelation(classForm)) {
av_relation* ar_entry = NULL;
av_toastid_mainid* at_entry = NULL;
bool found = false;
* is a logic table and has no data in physical files corresponding
* to its relfilenode
*/
if (!isPartitionedRelation(classForm)) {
at_entry =
(av_toastid_mainid*)hash_search(toast_table_map, &(classForm->reltoastrelid), HASH_ENTER, &found);
if (!found) {
at_entry->at_relid = relid;
at_entry->at_parentid = InvalidOid;
at_entry->at_allowvacuum = enable_vacuum;
at_entry->at_doanalyze = doanalyze;
at_entry->at_dovacuum = dovacuum;
at_entry->at_needfreeze = need_freeze;
at_entry->at_internal = is_internal_relation;
}
Assert(OidIsValid(at_entry->at_relid));
}
* Because of the design of partition toast table,
* we have to use to hash table to get reloptions for toast.
* One is below, save relations of oid and reloptions;
* The other is toastid-relid map, we get it from pg_class and pg_partition.
* !!!Unlike PG, we pass relid instand of toastid to the hash table.
*/
ar_entry = (av_relation*)hash_search(table_relopt_map, &relid, HASH_ENTER, &found);
if (!found) {
ar_entry->ar_hasrelopts = false;
if (relopts != NULL) {
ar_entry->ar_hasrelopts = true;
rc = memcpy_s(&ar_entry->ar_reloptions, sizeof(AutoVacOpts), relopts, sizeof(AutoVacOpts));
securec_check(rc, "", "");
}
}
}
if (relopts != NULL) {
pfree_ext(relopts);
}
}
tableam_scan_end(relScan);
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: Scan pg_class to determine which tables to vacuum");
* On the second pass, to collect all the partitions in the pg_partition,
* and also the partitioned table relid to TOAST relid mapping.
*/
ScanKeyInit(&key[0], Anum_pg_partition_parttype, BTEqualStrategyNumber, F_CHAREQ,
CharGetDatum(PART_OBJ_TYPE_TABLE_PARTITION));
partRel = heap_open(PartitionRelationId, AccessShareLock);
partScan = tableam_scan_begin(partRel, SnapshotNow, 1, key);
while (NULL != (partTuple = (HeapTuple) tableam_scan_getnexttuple(partScan, ForwardScanDirection))) {
Form_pg_partition partForm = (Form_pg_partition)GETSTRUCT(partTuple);
* vacuum the subpartition later. */
if (!OidIsValid(partForm->relfilenode)) {
continue;
}
PgStat_StatTabEntry* tabentry = NULL;
AutoVacOpts* relopts = NULL;
bool dovacuum = false;
bool doanalyze = false;
bool need_freeze = false;
Oid partOid;
bool found = false;
av_relation* ar_hentry = NULL;
at_partitioned_table* ap_entry = NULL;
* 'found = false' means partitioned table do autovac on other coordiantor
* coordiantor that analyze table partition is consistent with the coordiantor
* that analyze partition table.
*/
ar_hentry = (av_relation*)hash_search(table_relopt_map, &(partForm->parentid), HASH_FIND, &found);
if (!found)
continue;
if (ar_hentry->ar_hasrelopts)
relopts = &ar_hentry->ar_reloptions;
ap_entry = (at_partitioned_table*)hash_search(partitioned_tables_map, &partForm->parentid, HASH_FIND, &found);
if (!found) {
ereport(defence_errlevel(), (errcode(ERRCODE_DATA_CORRUPTED), errmsg("Oid: %u does not "
"find in partitioned tables map.", partForm->parentid)));
}
partOid = HeapTupleGetOid(partTuple);
tabentry = get_pgstat_tabentry_relid(partOid, false, partForm->parentid, shared, dbentry);
partition_needs_vacanalyze(partOid, relopts, partForm, partTuple, ap_entry, tabentry,
false, &dovacuum, &doanalyze, &need_freeze);
Assert(false == doanalyze);
if (freeze_autovacuum) {
dovacuum = need_freeze;
}
if (dovacuum) {
vacObj = (vacuum_object*)palloc(sizeof(vacuum_object));
vacObj->tab_oid = partOid;
vacObj->parent_oid = partForm->parentid;
vacObj->dovacuum = dovacuum;
vacObj->dovacuum_toast = false;
vacObj->doanalyze = doanalyze;
vacObj->need_freeze = need_freeze;
vacObj->is_internal_relation = false;
vacObj->gpi_vacuumed = false;
vacObj->flags = VACFLG_SUB_PARTITION;
table_oids = lappend(table_oids, vacObj);
}
if (OidIsValid(partForm->reltoastrelid)) {
av_toastid_mainid* at_entry = NULL;
at_entry = (av_toastid_mainid*)hash_search(toast_table_map, &(partForm->reltoastrelid), HASH_ENTER, &found);
if (!found) {
at_entry->at_relid = partOid;
at_entry->at_parentid = partForm->parentid;
at_entry->at_allowvacuum = ap_entry->at_allowvacuum;
at_entry->at_doanalyze = doanalyze | ap_entry->at_doanalyze;
at_entry->at_dovacuum = dovacuum | ap_entry->at_dovacuum;
at_entry->at_needfreeze = need_freeze;
at_entry->at_internal = false;
}
* if we found map but parentid doesn't equal to partFrom->parentid
* may be the reltoastrelid has been exchanged by some one,
* (e.g. alter table exchange partition)
* just skip it this time.
*/
if (found && (at_entry->at_parentid != partForm->parentid)) {
if (hash_search(toast_table_map, &(partForm->reltoastrelid), HASH_REMOVE, NULL) != NULL) {
ereport(LOG, (errmsg("reltoastrelid: %u toast table map "
"has been changed, skip it.", partForm->reltoastrelid)));
} else {
ereport(ERROR, (errcode(ERRCODE_DATA_CORRUPTED), errmsg("toast table map hash table corrupted.")));
}
} else {
Assert(OidIsValid(at_entry->at_relid) && OidIsValid(at_entry->at_parentid));
}
}
}
tableam_scan_end(partScan);
heap_close(partRel, AccessShareLock);
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: Scan pg_partition to determine which partitions to vacuum");
* Meanwhile, to collect all the subpartitions in the pg_partition,
* and also the partitioned table relid to TOAST relid mapping.
*/
ScanKeyInit(&key[0], Anum_pg_partition_parttype, BTEqualStrategyNumber, F_CHAREQ,
CharGetDatum(PART_OBJ_TYPE_TABLE_SUB_PARTITION));
partRel = heap_open(PartitionRelationId, AccessShareLock);
subpartScan = tableam_scan_begin(partRel, SnapshotNow, 1, key);
while (NULL != (partTuple = (HeapTuple) tableam_scan_getnexttuple(subpartScan, ForwardScanDirection))) {
Form_pg_partition partForm = (Form_pg_partition)GETSTRUCT(partTuple);
PgStat_StatTabEntry* tabentry = NULL;
AutoVacOpts* relopts = NULL;
bool dovacuum = false;
bool doanalyze = false;
bool need_freeze = false;
Oid partOid;
Oid tableOid;
bool found = false;
av_relation* ar_hentry = NULL;
at_partitioned_table* ap_entry = NULL;
tableOid = partid_get_parentid(partForm->parentid);
if (!OidIsValid(tableOid)) {
continue;
}
* 'found = false' means partitioned table do autovac on other coordiantor
* coordiantor that analyze table partition is consistent with the coordiantor
* that analyze partition table.
*/
ar_hentry = (av_relation*)hash_search(table_relopt_map, &tableOid, HASH_FIND, &found);
if (!found)
continue;
if (ar_hentry->ar_hasrelopts)
relopts = &ar_hentry->ar_reloptions;
ap_entry = (at_partitioned_table*)hash_search(partitioned_tables_map, &tableOid, HASH_FIND, &found);
if (!found) {
ereport(defence_errlevel(), (errcode(ERRCODE_DATA_CORRUPTED), errmsg("Oid: %u does not "
"find in partitioned tables map.", tableOid)));
}
partOid = HeapTupleGetOid(partTuple);
tabentry = get_pgstat_tabentry_relid(partOid, false, tableOid, shared, dbentry);
partition_needs_vacanalyze(partOid, relopts, partForm, partTuple, ap_entry, tabentry,
false, &dovacuum, &doanalyze, &need_freeze);
Assert(false == doanalyze);
if (freeze_autovacuum) {
dovacuum = need_freeze;
}
if (dovacuum) {
vacObj = (vacuum_object*)palloc(sizeof(vacuum_object));
vacObj->tab_oid = partOid;
vacObj->parent_oid = partForm->parentid;
vacObj->dovacuum = dovacuum;
vacObj->dovacuum_toast = false;
vacObj->doanalyze = doanalyze;
vacObj->need_freeze = need_freeze;
vacObj->is_internal_relation = false;
vacObj->gpi_vacuumed = false;
vacObj->flags = VACFLG_SUB_PARTITION;
table_oids = lappend(table_oids, vacObj);
}
if (OidIsValid(partForm->reltoastrelid)) {
av_toastid_mainid* at_entry = NULL;
at_entry = (av_toastid_mainid*)hash_search(toast_table_map, &(partForm->reltoastrelid), HASH_ENTER, &found);
if (!found) {
at_entry->at_relid = partOid;
at_entry->at_parentid = partForm->parentid;
at_entry->at_allowvacuum = ap_entry->at_allowvacuum;
at_entry->at_doanalyze = doanalyze | ap_entry->at_doanalyze;
at_entry->at_dovacuum = dovacuum | ap_entry->at_dovacuum;
at_entry->at_needfreeze = need_freeze;
at_entry->at_internal = false;
}
* if we found map but parentid doesn't equal to partFrom->parentid
* may be the reltoastrelid has been exchanged by some one,
* (e.g. alter table exchange partition)
* just skip it this time.
*/
if (found && (at_entry->at_parentid != partForm->parentid)) {
if (hash_search(toast_table_map, &(partForm->reltoastrelid), HASH_REMOVE, NULL) != NULL) {
ereport(LOG, (errmsg("reltoastrelid: %u toast table map "
"has been changed, skip it.", partForm->reltoastrelid)));
} else {
ereport(ERROR, (errcode(ERRCODE_DATA_CORRUPTED), errmsg("toast table map hash table corrupted.")));
}
} else {
Assert(OidIsValid(at_entry->at_relid) && OidIsValid(at_entry->at_parentid));
}
}
}
tableam_scan_end(subpartScan);
heap_close(partRel, AccessShareLock);
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: Scan pg_partition to determine which subpartitions to vacuum");
ScanKeyInit(&key[0], Anum_pg_class_relkind, BTEqualStrategyNumber, F_CHAREQ, CharGetDatum(RELKIND_TOASTVALUE));
relScan = tableam_scan_begin(classRel, SnapshotNow, 1, &key[0]);
while ((tuple = (HeapTuple) tableam_scan_getnexttuple(relScan, ForwardScanDirection)) != NULL) {
Form_pg_class classForm = (Form_pg_class)GETSTRUCT(tuple);
Oid relid = HeapTupleGetOid(tuple);
PgStat_StatTabEntry* tabentry = NULL;
AutoVacOpts* relopts = NULL;
bytea *rawRelopts = NULL;
bool isReloptsReferenceOther = false;
bool dovacuum = false;
bool doanalyze = false;
bool need_freeze = false;
bool enable_analyze = false;
bool enable_vacuum = false;
bool found = false;
av_toastid_mainid* at_entry = NULL;
if (classForm->relpersistence == RELPERSISTENCE_TEMP ||
classForm->relpersistence == RELPERSISTENCE_GLOBAL_TEMP)
continue;
at_entry = (av_toastid_mainid*)hash_search(toast_table_map, &(relid), HASH_FIND, &found);
* For Global Autovaccum
* 1. skip check if fail to find main table/partition
* 2. skip check if main table/partition have been taken be vaccumed
*/
if (!local_autovacuum && (!found || at_entry->at_dovacuum))
continue;
* fetch reloptions -- if this toast table does not have them, try the
* main rel
*/
relopts = extract_autovac_opts(tuple, pg_class_desc);
rawRelopts = extractRelOptions(tuple, pg_class_desc, InvalidOid);
* we must get main table id first, and then get the
* reloptions according to the main table id
*/
if (NULL == relopts && NULL != at_entry) {
av_relation* ar_hentry = NULL;
Oid MainId = at_entry->at_parentid > InvalidOid ? at_entry->at_parentid : at_entry->at_relid;
ar_hentry = (av_relation*)hash_search(table_relopt_map, &MainId, HASH_FIND, &found);
if (found && ar_hentry->ar_hasrelopts) {
relopts = &ar_hentry->ar_reloptions;
isReloptsReferenceOther = true;
}
}
tabentry = get_pgstat_tabentry_relid(relid, classForm->relisshared, InvalidOid, shared, dbentry);
relation_support_autoavac(tuple, &enable_analyze, &enable_vacuum, &is_internal_relation);
relation_needs_vacanalyze(relid, relopts, rawRelopts, classForm, tuple, tabentry, enable_analyze,
enable_vacuum, true, &dovacuum, &doanalyze, &need_freeze);
if (freeze_autovacuum) {
if (ISMATMAP(classForm->relname.data) || ISMLOG(classForm->relname.data)) {
dovacuum = true;
doanalyze = false;
} else {
dovacuum = need_freeze;
doanalyze = false;
}
}
if (dovacuum) {
vacObj = (vacuum_object*)palloc(sizeof(vacuum_object));
if (local_autovacuum) {
vacObj->tab_oid = relid;
vacObj->parent_oid = InvalidOid;
vacObj->flags = VACFLG_SIMPLE_HEAP;
vacObj->dovacuum_toast = false;
} else {
vacObj->tab_oid = at_entry->at_relid;
vacObj->parent_oid = at_entry->at_parentid;
vacObj->flags = OidIsValid(at_entry->at_parentid) ? VACFLG_SUB_PARTITION : VACFLG_SIMPLE_HEAP;
vacObj->dovacuum_toast = true;
vacObj->is_internal_relation = at_entry->at_internal;
}
vacObj->dovacuum = dovacuum;
vacObj->doanalyze = doanalyze;
vacObj->need_freeze = need_freeze;
table_oids = lappend(table_oids, vacObj);
}
if (relopts && !isReloptsReferenceOther) {
pfree_ext(relopts);
}
}
tableam_scan_end(relScan);
heap_close(classRel, AccessShareLock);
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: Scan pg_class to determine which toast tables to vacuum");
* Create one buffer access strategy object per buffer pool for VACUUM to use.
* We want to use the same one across all the vacuum operations we perform,
* since the point is for VACUUM not to blow out the shared cache.
*/
bstrategy = GetAccessStrategy(BAS_VACUUM);
* create a memory context to act as fake t_thrd.mem_cxt.portal_mem_cxt, so that the
* contexts created in the vacuum code are cleaned up for each table.
*/
t_thrd.mem_cxt.portal_mem_cxt = AllocSetContextCreate(t_thrd.autovacuum_cxt.AutovacMemCxt, "Autovacuum Portal",
ALLOCSET_DEFAULT_INITSIZE, ALLOCSET_DEFAULT_MINSIZE, ALLOCSET_DEFAULT_MAXSIZE);
* Perform operations on collected tables.
*/
foreach (cell, table_oids) {
Oid relid;
Oid parentid;
autovac_table* tab = NULL;
WorkerInfo worker = NULL;
bool skipit = false;
bool found = false;
int stdVacuumCostDelay;
int stdVacuumCostLimit;
at_partitioned_table* ap_entry = NULL;
vacObj = (vacuum_object*)lfirst(cell);
relid = vacObj->tab_oid;
parentid = vacObj->parent_oid;
if (!u_sess->attr.attr_storage.autovacuum_start_daemon && !vacObj->need_freeze)
break;
CHECK_FOR_INTERRUPTS();
* hold schedule lock from here until we're sure that this table still
* needs vacuuming. We also need the AutovacuumLock to walk the
* worker array, but we'll let go of that one quickly.
*/
LWLockAcquire(AutovacuumScheduleLock, LW_EXCLUSIVE);
LWLockAcquire(AutovacuumLock, LW_SHARED);
* Check whether the table is being vacuumed concurrently by another
* worker.
*/
skipit = false;
worker = (WorkerInfo)SHMQueueNext(&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers,
&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers, offsetof(WorkerInfoData, wi_links));
while (worker != NULL) {
if (worker == t_thrd.autovacuum_cxt.MyWorkerInfo)
goto next_worker;
if (!worker->wi_sharedrel && worker->wi_dboid != u_sess->proc_cxt.MyDatabaseId)
goto next_worker;
* we can not identify it only by oid.
* check the main table:
* 1. other worker handle the main table, need check the worker's tableoid not equal the relid;
* 2. other worker handle the part table, need check the worker's parentoid not equal the relid;
* check the part table:
* 1. other worker handle the main table, need check the worker's tableoid not equal the parentid;
* 2. other worker handle the part table, need check the worker's parentoid not equal the parentid;
*/
if (parentid == InvalidOid && (worker->wi_tableoid == relid || worker->wi_parentoid == relid)) {
AUTOVAC_LOG(DEBUG1, "parentoid = %u, tableoid = %u is on autovac, just skip it", parentid, relid);
skipit = true;
break;
}
if (parentid != InvalidOid && (worker->wi_tableoid == parentid || worker->wi_parentoid == parentid)) {
AUTOVAC_LOG(DEBUG1, "parentoid = %u, tableoid = %u is on autovac, just skip it", parentid, relid);
skipit = true;
break;
}
next_worker:
worker = (WorkerInfo)SHMQueueNext(&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers,
&worker->wi_links, offsetof(WorkerInfoData, wi_links));
}
LWLockRelease(AutovacuumLock);
if (skipit) {
LWLockRelease(AutovacuumScheduleLock);
continue;
}
* Check whether pgstat data still says we need to vacuum this table.
* It could have changed if something else processed the table while
* we weren't looking.
*
* Note: we have a special case in pgstat code to ensure that the
* stats we read are as up-to-date as possible, to avoid the problem
* that somebody just finished vacuuming this table. The window to
* the race condition is not closed but it is very small.
*/
(void)MemoryContextSwitchTo(t_thrd.autovacuum_cxt.AutovacMemCxt);
if ((vacObj->flags & VACFLG_SIMPLE_HEAP) || (vacObj->flags & VACFLG_MAIN_PARTITION)) {
tab = table_recheck_autovac(vacObj, table_relopt_map, toast_table_map, pg_class_desc);
} else {
Assert(vacObj->flags & VACFLG_SUB_PARTITION);
tab = partition_recheck_autovac(vacObj, table_relopt_map, partitioned_tables_map, pg_class_desc);
}
if (tab == NULL) {
LWLockRelease(AutovacuumScheduleLock);
continue;
}
tab->at_flags = vacObj->flags;
* Ok, good to go. Store the table in shared memory before releasing
* the lock so that other workers don't vacuum it concurrently.
*/
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_tableoid = relid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_parentoid = parentid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_sharedrel = tab->at_sharedrel;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_ispartition = vacuumPartition((uint32)(vacObj->flags));
LWLockRelease(AutovacuumScheduleLock);
* Remember the prevailing values of the vacuum cost GUCs. We have to
* restore these at the bottom of the loop, else we'll compute wrong
* values in the next iteration of autovac_balance_cost().
*/
stdVacuumCostDelay = u_sess->attr.attr_storage.VacuumCostDelay;
stdVacuumCostLimit = u_sess->attr.attr_storage.VacuumCostLimit;
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_delay = tab->at_vacuum_cost_delay;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_limit = tab->at_vacuum_cost_limit;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_cost_limit_base = tab->at_vacuum_cost_limit;
autovac_balance_cost();
AutoVacuumUpdateDelay();
LWLockRelease(AutovacuumLock);
MemoryContextResetAndDeleteChildren(t_thrd.mem_cxt.portal_mem_cxt);
(void)MemoryContextSwitchTo(t_thrd.mem_cxt.portal_mem_cxt);
* Save the relation name for a possible error message, to avoid a
* catalog lookup in case of an error. If any of these return NULL,
* then the relation has been dropped since last we checked; skip it.
* Note: they must live in a long-lived memory context because we call
* vacuum and analyze in different transactions.
*/
if (vacuumPartition((uint32)(vacObj->flags))) {
Oid at_parentid = partid_get_parentid(tab->at_relid);
Oid at_grandparentid = partid_get_parentid(at_parentid);
if (OidIsValid(at_grandparentid)) {
tab->at_subpartname = getPartitionName(tab->at_relid, false);
tab->at_partname = NULL;
tab->at_relname = get_rel_name(at_grandparentid);
tab->at_nspname = get_namespace_name(get_rel_namespace(at_grandparentid));
} else {
tab->at_subpartname = NULL;
tab->at_partname = getPartitionName(tab->at_relid, false);
tab->at_relname = get_rel_name(at_parentid);
tab->at_nspname = get_namespace_name(get_rel_namespace(at_parentid));
}
} else {
tab->at_subpartname = NULL;
tab->at_partname = NULL;
tab->at_relname = get_rel_name(tab->at_relid);
tab->at_nspname = get_namespace_name(get_rel_namespace(tab->at_relid));
}
tab->at_datname = get_database_name(u_sess->proc_cxt.MyDatabaseId);
if ((NULL == tab->at_relname) || (NULL == tab->at_nspname) || (NULL == tab->at_datname))
goto deleted;
* We will abort vacuuming the current table if something errors out,
* and continue with the next one in schedule; in particular, this
* happens if we are interrupted with SIGINT.
*/
PG_TRY();
{
* 1. Let pgstat know what we're doing
* 2. in this function, statement_timestamp will be set to current time
*/
autovac_report_activity(tab);
if (ActiveSnapshotSet())
PopActiveSnapshot();
CommitTransactionCommand();
StartTransactionCommand();
LockSharedObject(DatabaseRelationId, u_sess->proc_cxt.MyDatabaseId, 0, RowExclusiveLock);
PushActiveSnapshot(GetTransactionSnapshot());
(void)MemoryContextSwitchTo(t_thrd.mem_cxt.portal_mem_cxt);
if ((IS_PGXC_COORDINATOR || IS_SINGLE_NODE) && u_sess->attr.attr_storage.autoanalyze_timeout > 0 &&
tab->at_doanalyze && !tab->at_dovacuum) {
t_thrd.storage_cxt.timeIsPausing = false;
enable_sig_alarm(u_sess->attr.attr_storage.autoanalyze_timeout * 1000, true);
}
if (!check_table_is_locked(tab)) {
if (local_autovacuum || vacObj->is_internal_relation)
autovacuum_local_vac_analyze(tab, bstrategy);
else
autovacuum_do_vac_analyze(tab, bstrategy);
}
if (vacObj->flags & VACFLG_SUB_PARTITION) {
Oid table_oid = parentid;
Oid grandparentid = partid_get_parentid(parentid);
if (OidIsValid(grandparentid)) {
table_oid = grandparentid;
}
ap_entry = (at_partitioned_table*)hash_search(partitioned_tables_map, &table_oid, HASH_FIND, &found);
if (found && !ap_entry->at_gpivacuumed && tab->at_gpivacuumed) {
ap_entry->at_gpivacuumed = tab->at_gpivacuumed;
}
}
disable_sig_alarm(true);
if (t_thrd.proc->workingVersionNum >= STATISTIC_HISTORY_VERSION_NUMBER
&& u_sess->attr.attr_sql.gms_stats_history_retention != -1) {
ClearStatsHistoryTimeoutRecords();
}
* Clear a possible query-cancel signal, to avoid a late reaction
* to an automatically-sent signal because of vacuuming the
* current table (we're done with it, so it would make no sense to
* cancel at this point.)
*/
t_thrd.int_cxt.QueryCancelPending = false;
}
PG_CATCH();
{
bool timeout_flag =
(t_thrd.storage_cxt.cancel_from_timeout && u_sess->attr.attr_storage.autoanalyze_timeout);
* Abort the transaction, start a new one, and proceed with the
* next table in our list.
*/
HOLD_INTERRUPTS();
t_thrd.int_cxt.QueryCancelPending = false;
(void)disable_sig_alarm(true);
t_thrd.int_cxt.QueryCancelPending = false;
t_thrd.storage_cxt.timeIsPausing = false;
t_thrd.storage_cxt.restimems = -1;
if (tab->at_dovacuum)
errcontext("automatic vacuum of table \"%s.%s.%s\"", tab->at_datname, tab->at_nspname, tab->at_relname);
else
errcontext(
"automatic analyze of table \"%s.%s.%s\"", tab->at_datname, tab->at_nspname, tab->at_relname);
EmitErrorReport();
AbortCurrentTransaction();
FlushErrorState();
MemoryContextResetAndDeleteChildren(t_thrd.mem_cxt.msg_mem_cxt);
MemoryContextResetAndDeleteChildren(t_thrd.mem_cxt.portal_mem_cxt);
gs_signal_setmask(&t_thrd.libpq_cxt.UnBlockSig, NULL);
(void)gs_signal_unblock_sigusr2();
StartTransactionCommand();
if (timeout_flag) {
Oid grandparent_oid = partid_get_parentid(vacObj->parent_oid);
Oid statFlag = OidIsValid(grandparent_oid) ? grandparent_oid : vacObj->parent_oid;
pgstat_report_autovac_timeout(vacObj->tab_oid, statFlag, tab->at_sharedrel);
}
RESUME_INTERRUPTS();
LockSharedObject(DatabaseRelationId, u_sess->proc_cxt.MyDatabaseId, 0, RowExclusiveLock);
}
PG_END_TRY();
deleted:
* Remove my info from shared memory. We could, but intentionally
* don't, clear wi_cost_limit and friends --- this is on the
* assumption that we probably have more to do with similar cost
* settings, so we don't want to give up our share of I/O for a very
* short interval and thereby thrash the global balance.
*/
LWLockAcquire(AutovacuumLock, LW_EXCLUSIVE);
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_tableoid = InvalidOid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_parentoid = InvalidOid;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_sharedrel = false;
t_thrd.autovacuum_cxt.MyWorkerInfo->wi_ispartition = false;
LWLockRelease(AutovacuumLock);
u_sess->attr.attr_storage.VacuumCostDelay = stdVacuumCostDelay;
u_sess->attr.attr_storage.VacuumCostLimit = stdVacuumCostLimit;
MemoryContextResetAndDeleteChildren(t_thrd.mem_cxt.msg_mem_cxt);
MemoryContextResetAndDeleteChildren(t_thrd.mem_cxt.portal_mem_cxt);
t_thrd.vacuum_cxt.vac_context = NULL;
}
* We leak table_toast_map here (among other things), but since we're
* going away soon, it's not a problem.
*/
* Update pg_database.datfrozenxid, and truncate pg_clog if possible. We
* only need to do this once, not after each table.
*/
vac_update_datfrozenxid();
if (ActiveSnapshotSet())
PopActiveSnapshot();
CommitTransactionCommand();
}
* extract_autovac_opts
*
* Given a relation's pg_class tuple, return the AutoVacOpts portion of
* reloptions, if set; otherwise, return NULL.
*/
AutoVacOpts* extract_autovac_opts(HeapTuple tup, TupleDesc pg_class_desc)
{
bytea* relopts = NULL;
AutoVacOpts* av = NULL;
int rc = 0;
Assert(((Form_pg_class)GETSTRUCT(tup))->relkind == RELKIND_RELATION ||
((Form_pg_class) GETSTRUCT(tup))->relkind == RELKIND_MATVIEW ||
((Form_pg_class)GETSTRUCT(tup))->relkind == RELKIND_TOASTVALUE);
relopts = extractRelOptions(tup, pg_class_desc, InvalidOid);
if (relopts == NULL)
return NULL;
av = (AutoVacOpts*)palloc(sizeof(AutoVacOpts));
rc = memcpy_s(av, sizeof(AutoVacOpts), &(((StdRdOptions*)relopts)->autovacuum), sizeof(AutoVacOpts));
securec_check(rc, "\0", "\0");
pfree_ext(relopts);
return av;
}
* get_pgstat_tabentry_relid
*
* Fetch the pgstat entry of a table, either local to a database or shared.
*/
static PgStat_StatTabEntry* get_pgstat_tabentry_relid(
Oid relid, bool isshared, uint32 statFlag, PgStat_StatDBEntry* shared, PgStat_StatDBEntry* dbentry)
{
PgStat_StatTabEntry* tabentry = NULL;
PgStat_StatDBEntry* dnentry = NULL;
if (isshared)
dnentry = shared;
else
dnentry = dbentry;
if (dnentry != NULL) {
PgStat_StatTabKey tabkey;
tabkey.statFlag = statFlag;
tabkey.tableid = relid;
tabentry = (PgStat_StatTabEntry*)hash_search(dnentry->tables, (void*)(&tabkey), HASH_FIND, NULL);
}
return tabentry;
}
* calculate_vacuum_cost_and_freezeages
*
* Calculate the vacuum cost parameters and the freeze ages. If there
* are options set in pg_class.reloptions, use them; in the case of a
* toast table, try the main table too. Otherwise use the GUC
* defaults, autovacuum's own first and plain vacuum second.
*/
static autovac_table* calculate_vacuum_cost_and_freezeages(const AutoVacOpts* avopts, bool doanalyze, bool need_freeze)
{
int64 freeze_min_age;
int64 freeze_table_age;
int vac_cost_limit;
int vac_cost_delay;
autovac_table* tab = NULL;
vac_cost_delay = (avopts && avopts->vacuum_cost_delay >= 0)
? avopts->vacuum_cost_delay
: (u_sess->attr.attr_storage.autovacuum_vac_cost_delay >= 0)
? u_sess->attr.attr_storage.autovacuum_vac_cost_delay
: u_sess->attr.attr_storage.VacuumCostDelay;
vac_cost_limit = (avopts && avopts->vacuum_cost_limit > 0)
? avopts->vacuum_cost_limit
: (u_sess->attr.attr_storage.autovacuum_vac_cost_limit > 0)
? u_sess->attr.attr_storage.autovacuum_vac_cost_limit
: u_sess->attr.attr_storage.VacuumCostLimit;
freeze_min_age =
(avopts && avopts->freeze_min_age >= 0) ? avopts->freeze_min_age : t_thrd.autovacuum_cxt.default_freeze_min_age;
freeze_table_age = (avopts && avopts->freeze_table_age >= 0) ? avopts->freeze_table_age
: t_thrd.autovacuum_cxt.default_freeze_table_age;
tab = (autovac_table*)palloc(sizeof(autovac_table));
tab->at_doanalyze = doanalyze;
tab->at_freeze_min_age = freeze_min_age;
tab->at_freeze_table_age = freeze_table_age;
tab->at_vacuum_cost_limit = vac_cost_limit;
tab->at_vacuum_cost_delay = vac_cost_delay;
tab->at_needfreeze = need_freeze;
tab->at_relname = NULL;
tab->at_nspname = NULL;
tab->at_datname = NULL;
return tab;
}
* table_recheck_autovac
*
* Recheck whether a table still needs vacuum or analyze. Return value is a
* valid autovac_table pointer if it does, NULL otherwise.
*
* Note that the returned autovac_table does not have the name fields set.
*/
static autovac_table* table_recheck_autovac(
vacuum_object* vacObj, HTAB* table_relopt_map, HTAB* toast_table_map, TupleDesc pg_class_desc)
{
Oid relid = vacObj->tab_oid;
Form_pg_class classForm;
HeapTuple classTup;
bool dovacuum = false;
bool dovacuum_toast = vacObj->dovacuum_toast;
bool doanalyze = false;
bool need_freeze = false;
bool is_internal_relation = false;
bool enable_analyze = false;
bool enable_vacuum = false;
autovac_table* tab = NULL;
PgStat_StatDBEntry* shared = NULL;
PgStat_StatDBEntry* dbentry = NULL;
PgStat_StatTabEntry* tabentry = NULL;
AutoVacOpts* avopts = NULL;
bool isAvoptsRefereceOther = false;
if (IS_SINGLE_NODE) {
autovac_refresh_stats();
}
shared = pgstat_fetch_stat_dbentry(InvalidOid);
dbentry = pgstat_fetch_stat_dbentry(u_sess->proc_cxt.MyDatabaseId);
classTup = SearchSysCacheCopy1(RELOID, ObjectIdGetDatum(relid));
if (!HeapTupleIsValid(classTup))
return NULL;
classForm = (Form_pg_class)GETSTRUCT(classTup);
bytea *rawRelopts = extractRelOptions(classTup, pg_class_desc, InvalidOid);
bool isUstorePartitionTable = (rawRelopts != NULL && RelationIsTableAccessMethodUStoreType(rawRelopts) &&
isPartitionedRelation(classForm));
* Get the applicable reloptions. If it is a TOAST table, try to get the
* main table reloptions if the toast table itself doesn't have.
*/
avopts = extract_autovac_opts(classTup, pg_class_desc);
if (RELKIND_TOASTVALUE == classForm->relkind && (NULL == avopts) && (NULL != table_relopt_map) &&
(NULL != toast_table_map)) {
av_relation* hentry = NULL;
av_toastid_mainid* tentry = NULL;
bool found = false;
Oid MainId;
tentry = (av_toastid_mainid*)hash_search(toast_table_map, &relid, HASH_FIND, &found);
if (found) {
MainId = tentry->at_relid;
hentry = (av_relation*)hash_search(table_relopt_map, &MainId, HASH_FIND, &found);
if (found && hentry->ar_hasrelopts) {
avopts = &hentry->ar_reloptions;
isAvoptsRefereceOther = true;
}
}
}
tabentry = get_pgstat_tabentry_relid(relid, classForm->relisshared, InvalidOid, shared, dbentry);
relation_support_autoavac(classTup, &enable_analyze, &enable_vacuum, &is_internal_relation);
relation_needs_vacanalyze(relid, avopts, rawRelopts, classForm, classTup, tabentry, enable_analyze,
enable_vacuum, true, &dovacuum, &doanalyze, &need_freeze);
if (classForm->relkind == RELKIND_TOASTVALUE)
doanalyze = false;
if (ISMATMAP(classForm->relname.data) || ISMLOG(classForm->relname.data)) {
dovacuum = true;
doanalyze = true;
}
if (doanalyze || dovacuum || dovacuum_toast || isUstorePartitionTable) {
tab = calculate_vacuum_cost_and_freezeages(avopts, doanalyze, need_freeze);
if (tab != NULL) {
tab->at_relid = relid;
tab->at_sharedrel = classForm->relisshared;
tab->at_dovacuum = isUstorePartitionTable ? true : (dovacuum || dovacuum_toast);
tab->at_gpivacuumed = vacObj->gpi_vacuumed;
}
}
heap_freetuple(classTup);
if (avopts && !isAvoptsRefereceOther) {
pfree_ext(avopts);
}
return tab;
}
* determine_vacuum_params
* Determine vacuum/analyze equation parameters. We have two possible
* sources: the passed reloptions (which could be a main table or a toast
* table), or the autovacuum GUC variables.
*
*/
static void determine_vacuum_params(float4& vac_scale_factor, int& vac_base_thresh, float4& anl_scale_factor,
int& anl_base_thresh, int64& freeze_max_age, bool& av_enabled, TransactionId& xidForceLimit,
MultiXactId& multiForceLimit, const AutoVacOpts* relopts)
{
vac_scale_factor = (relopts && relopts->vacuum_scale_factor >= 0) ? relopts->vacuum_scale_factor
: u_sess->attr.attr_storage.autovacuum_vac_scale;
vac_base_thresh = (relopts && relopts->vacuum_threshold >= 0) ? relopts->vacuum_threshold
: u_sess->attr.attr_storage.autovacuum_vac_thresh;
anl_scale_factor = (relopts && relopts->analyze_scale_factor >= 0) ? relopts->analyze_scale_factor
: u_sess->attr.attr_storage.autovacuum_anl_scale;
anl_base_thresh = (relopts && relopts->analyze_threshold >= 0) ? relopts->analyze_threshold
: u_sess->attr.attr_storage.autovacuum_anl_thresh;
freeze_max_age = (relopts && relopts->freeze_max_age >= 0)
? Min(relopts->freeze_max_age, g_instance.attr.attr_storage.autovacuum_freeze_max_age)
: g_instance.attr.attr_storage.autovacuum_freeze_max_age;
av_enabled = (relopts ? relopts->enabled : true);
if (t_thrd.autovacuum_cxt.recentXid > FirstNormalTransactionId + freeze_max_age)
xidForceLimit = t_thrd.autovacuum_cxt.recentXid - freeze_max_age;
else
xidForceLimit = FirstNormalTransactionId;
#ifndef ENABLE_MULTIPLE_NODES
if (t_thrd.autovacuum_cxt.recentMulti >
FirstMultiXactId + (uint64)g_instance.attr.attr_storage.autovacuum_freeze_max_age)
multiForceLimit = t_thrd.autovacuum_cxt.recentMulti -
(uint64)g_instance.attr.attr_storage.autovacuum_freeze_max_age;
else
multiForceLimit = FirstMultiXactId;
#endif
}
* relation_needs_vacanalyze
*
* Check whether a relation needs to be vacuumed or analyzed; return each into
* "dovacuum" and "doanalyze", respectively. Also return whether the vacuum is
* being forced because need freeze the old tuple for recycle clog.
*
* relopts is a pointer to the AutoVacOpts options (either for itself in the
* case of a plain table, or for either itself or its parent table in the case
* of a TOAST table), NULL if none; tabentry is the pgstats entry, which can be
* NULL.
*
* A table needs to be vacuumed if the number of dead tuples exceeds a
* threshold. This threshold is calculated as
*
* threshold = vac_base_thresh + vac_scale_factor * reltuples
*
* For analyze, the analysis done is that the number of tuples inserted,
* deleted and updated since the last analyze exceeds a threshold calculated
* in the same fashion as above. Note that the collector actually stores
* the number of tuples (both live and dead) that there were as of the last
* analyze. This is asymmetric to the VACUUM case.
*
* We also force vacuum if the table's relfrozenxid64 is more than freeze_max_age
* transactions back.
*
* A table whose autovacuum_enabled option is false is
* automatically skipped (unless we have to vacuum it due to freeze_max_age).
* Thus autovacuum can be disabled for specific tables. Also, when the stats
* collector does not have data about a table, it will be skipped.
*
* A table whose vac_base_thresh value is < 0 takes the base value from the
* autovacuum_vacuum_threshold GUC variable. Similarly, a vac_scale_factor
* value < 0 is substituted with the value of
* autovacuum_vacuum_scale_factor GUC variable. Ditto for analyze.
*/
static void relation_needs_vacanalyze(Oid relid, AutoVacOpts* relopts, bytea* rawRelopts, Form_pg_class classForm, HeapTuple tuple,
PgStat_StatTabEntry* tabentry, bool allowAnalyze, bool allowVacuum, bool is_recheck,
bool* dovacuum, bool* doanalyze, bool* need_freeze)
{
PgStat_StatTabKey tablekey;
avw_info* avwentry = NULL;
bool found = false;
bool force_vacuum = false;
bool delta_vacuum = false;
bool av_enabled = false;
bool userEnabled = true;
float4 reltuples;
int vac_base_thresh = 0;
int anl_base_thresh = 0;
float4 vac_scale_factor = 0.0;
float4 anl_scale_factor = 0.0;
float4 vacthresh;
float4 anlthresh;
int64 vactuples = 0;
int64 anltuples = 0;
int64 freeze_max_age = 0;
TransactionId xidForceLimit = InvalidTransactionId;
MultiXactId multiForceLimit = InvalidMultiXactId;
AssertArg(classForm != NULL);
AssertArg(OidIsValid(relid));
determine_vacuum_params(vac_scale_factor, vac_base_thresh, anl_scale_factor, anl_base_thresh, freeze_max_age,
av_enabled, xidForceLimit, multiForceLimit, relopts);
bool isNull = false;
TransactionId relfrozenxid = InvalidTransactionId;
Relation rel = heap_open(RelationRelationId, AccessShareLock);
Datum xid64datum = heap_getattr(tuple, Anum_pg_class_relfrozenxid64, RelationGetDescr(rel), &isNull);
if (isNull) {
relfrozenxid = classForm->relfrozenxid;
if (TransactionIdPrecedes(t_thrd.xact_cxt.ShmemVariableCache->nextXid, relfrozenxid) ||
!TransactionIdIsNormal(relfrozenxid)) {
relfrozenxid = FirstNormalTransactionId;
}
} else {
relfrozenxid = DatumGetTransactionId(xid64datum);
}
force_vacuum = (TransactionIdIsNormal(relfrozenxid) && TransactionIdPrecedes(relfrozenxid, xidForceLimit));
#ifndef ENABLE_MULTIPLE_NODES
if (!force_vacuum) {
Datum mxidDatum = heap_getattr(tuple, Anum_pg_class_relminmxid, RelationGetDescr(rel), &isNull);
MultiXactId relminmxid = isNull ? FirstMultiXactId : DatumGetTransactionId(mxidDatum);
force_vacuum = (MultiXactIdIsValid(relminmxid) && MultiXactIdPrecedes(relminmxid, multiForceLimit));
}
#endif
heap_close(rel, AccessShareLock);
*need_freeze = force_vacuum;
AUTOVAC_LOG(DEBUG2, "vac \"%s\": need freeze is %s", NameStr(classForm->relname), force_vacuum ? "true" : "false");
if (rawRelopts != NULL && StdRelOptIsColStore(rawRelopts) &&
g_instance.attr.attr_storage.enable_delta_store && DO_VACUUM) {
PgStat_StatDBEntry *dbentry = pgstat_fetch_stat_dbentry(u_sess->proc_cxt.MyDatabaseId);;
PgStat_StatDBEntry *shared = pgstat_fetch_stat_dbentry(InvalidOid);
PgStat_StatTabEntry *deltaTabentry = get_pgstat_tabentry_relid(classForm->reldeltarelid,
classForm->relisshared, InvalidOid, shared, dbentry);
if (deltaTabentry != NULL) {
delta_vacuum = (deltaTabentry->n_live_tuples >= ((StdRdOptions*)rawRelopts)->delta_rows_threshold);
}
}
if (!force_vacuum && (!av_enabled || !u_sess->attr.attr_storage.autovacuum_start_daemon)) {
userEnabled = false;
}
if (NULL != t_thrd.autovacuum_cxt.pgStatAutoVacInfo) {
tablekey.statFlag = InvalidOid;
tablekey.tableid = relid;
avwentry =
(avw_info*)hash_search(t_thrd.autovacuum_cxt.pgStatAutoVacInfo, (void*)(&tablekey), HASH_FIND, &found);
}
reltuples = classForm->reltuples;
vacthresh = (float4)vac_base_thresh + vac_scale_factor * reltuples;
anlthresh = (float4)anl_base_thresh + anl_scale_factor * reltuples;
if ((avwentry == NULL) && (tabentry == NULL)) {
*dovacuum = force_vacuum;
*doanalyze = false;
} else {
if (tabentry && (tabentry->changes_since_analyze || tabentry->n_dead_tuples)) {
anltuples = tabentry->changes_since_analyze;
vactuples = tabentry->n_dead_tuples;
AUTOVAC_LOG(DEBUG2, "fetch local stat info: vac \"%s\" changes_since_analyze = %ld n_dead_tuples = %ld ",
NameStr(classForm->relname), tabentry->changes_since_analyze, tabentry->n_dead_tuples);
}
if (avwentry && (avwentry->changes_since_analyze || avwentry->n_dead_tuples)) {
anltuples = avwentry->changes_since_analyze;
vactuples = avwentry->n_dead_tuples;
AUTOVAC_LOG(DEBUG2, "fetch global stat info: vac \"%s\" changes_since_analyze = %ld n_dead_tuples = %ld ",
NameStr(classForm->relname), avwentry->changes_since_analyze, avwentry->n_dead_tuples);
}
*dovacuum = force_vacuum || delta_vacuum;
*doanalyze = false;
if (false == *dovacuum && allowVacuum)
*dovacuum = ((float4)vactuples > vacthresh);
if (allowAnalyze)
*doanalyze = ((float4)anltuples > anlthresh);
}
*dovacuum = *dovacuum && userEnabled;
if (*dovacuum || *doanalyze) {
AUTOVAC_LOG(DEBUG2, "vac \"%s\": recheck = %s need_freeze = %s dovacuum = %s (dead tuples %ld "
"vacuum threshold %.0f) doanalyze = %s (changed tuples %ld analyze threshold %.0f)",
NameStr(classForm->relname), is_recheck ? "true" : "false", *need_freeze ? "true" : "false",
*dovacuum ? "true" : "false", vactuples, vacthresh, *doanalyze ? "true" : "false", anltuples, anlthresh);
}
DEBUG_VACUUM_LOG(relid, classForm->relnamespace, LOG, "vac \"%s\": recheck = %s need_freeze = %s dovacuum = %s "
"(dead tuples %ld vacuum threshold %.0f) doanalyze = %s (changed tuples %ld analyze threshold %.0f) reltuples "
"= %.0f", NameStr(classForm->relname), is_recheck ? "true" : "false", *need_freeze ? "true" : "false",
*dovacuum ? "true" : "false", vactuples, vacthresh, *doanalyze ? "true" : "false",
anltuples, anlthresh, reltuples);
}
* fill_in_vac_stmt
*
* fill in the vacuum statement.
*/
static void fill_in_vac_stmt(VacuumStmt& vacstmt, const autovac_table& tab, RangeVar* rangevar)
{
vacstmt.type = T_VacuumStmt;
if (!tab.at_needfreeze)
vacstmt.options = VACOPT_NOWAIT;
if (tab.at_dovacuum)
vacstmt.options = (unsigned int)vacstmt.options | VACOPT_VACUUM;
if (tab.at_doanalyze)
vacstmt.options = (unsigned int)vacstmt.options | VACOPT_ANALYZE;
#ifdef ENABLE_MOT
vacstmt.options |= VACOPT_AUTOVAC;
#endif
vacstmt.flags = tab.at_flags;
vacstmt.rely_oid = InvalidOid;
vacstmt.freeze_min_age = tab.at_freeze_min_age;
vacstmt.freeze_table_age = tab.at_freeze_table_age;
vacstmt.relation = rangevar;
vacstmt.va_cols = NIL;
vacstmt.gpi_vacuumed = tab.at_gpivacuumed;
vacstmt.needFreeze = tab.at_needfreeze;
}
* autovacuum_do_vac_analyze
* Vacuum and/or analyze the specified table
*/
static void autovacuum_do_vac_analyze(autovac_table* tab, BufferAccessStrategy bstrategy)
{
VacuumStmt vacstmt;
RangeVar rangevar;
const char* nspname = NULL;
const char* relname = NULL;
StringInfoData str;
errno_t rc = EOK;
rc = memset_s(&vacstmt, sizeof(vacstmt), 0, sizeof(vacstmt));
securec_check(rc, "", "");
rc = memset_s(&rangevar, sizeof(rangevar), 0, sizeof(rangevar));
securec_check(rc, "", "");
rangevar.schemaname = tab->at_nspname;
rangevar.relname = tab->at_relname;
if (NULL != tab->at_partname) {
rangevar.ispartition = true;
rangevar.partitionname = tab->at_partname;
} else if (NULL != tab->at_subpartname) {
rangevar.issubpartition = true;
rangevar.subpartitionname = tab->at_subpartname;
}
rangevar.location = -1;
nspname = quote_identifier(tab->at_nspname);
relname = quote_identifier(tab->at_relname);
fill_in_vac_stmt(vacstmt, *tab, &rangevar);
initStringInfo(&str);
if (tab->at_dovacuum)
appendStringInfo(&str, "VACUUM ");
if (tab->at_doanalyze)
appendStringInfo(&str, "ANALYZE ");
appendStringInfo(&str, "%s.%s", nspname, relname);
if (NULL != tab->at_partname) {
appendStringInfo(&str, " PARTITION (%s)", quote_identifier(tab->at_partname));
} else if (NULL != tab->at_subpartname) {
appendStringInfo(&str, " SUBPARTITION (%s)", quote_identifier(tab->at_subpartname));
}
WaitStatePhase oldPhase = pgstat_report_waitstatus_phase(PHASE_AUTOVACUUM);
DEBUG_MOD_START_TIMER(MOD_AUTOVAC);
DoVacuumMppTable(&vacstmt, str.data, true, false);
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: %s", str.data);
pgstat_report_waitstatus_phase(oldPhase);
}
* autovacuum_do_vac_analyze
* Vacuum and/or analyze the specified table
*/
static void autovacuum_local_vac_analyze(autovac_table* tab, BufferAccessStrategy bstrategy)
{
VacuumStmt vacstmt;
RangeVar rangevar;
errno_t rc = EOK;
rc = memset_s(&vacstmt, sizeof(vacstmt), 0, sizeof(vacstmt));
securec_check(rc, "", "");
rc = memset_s(&rangevar, sizeof(rangevar), 0, sizeof(rangevar));
securec_check(rc, "", "");
rangevar.schemaname = tab->at_nspname;
rangevar.relname = tab->at_relname;
rangevar.partitionname = tab->at_partname;
rangevar.subpartitionname = tab->at_subpartname;
rangevar.location = -1;
fill_in_vac_stmt(vacstmt, *tab, &rangevar);
autovac_report_activity(tab);
WaitStatePhase oldPhase = pgstat_report_waitstatus_phase(PHASE_AUTOVACUUM);
DEBUG_MOD_START_TIMER(MOD_AUTOVAC);
vacuum(&vacstmt, tab->at_relid, u_sess->attr.attr_storage.handle_toast_in_autovac, bstrategy, true);
tab->at_gpivacuumed = vacstmt.gpi_vacuumed;
DEBUG_MOD_STOP_TIMER(MOD_AUTOVAC, "AUTOVAC TIMER: %s", tab->at_relname);
pgstat_report_waitstatus_phase(oldPhase);
}
* autovac_report_activity
* Report to pgstat what autovacuum is doing
*
* We send a SQL string corresponding to what the user would see if the
* equivalent command was to be issued manually.
*
* Note we assume that we are going to report the next command as soon as we're
* done with the current one, and exit right after the last one, so we don't
* bother to report "<IDLE>" or some such.
*/
static void autovac_report_activity(autovac_table* tab)
{
#define MAX_AUTOVAC_ACTIV_LEN (NAMEDATALEN * 2 + 56)
char activity[MAX_AUTOVAC_ACTIV_LEN];
int len;
int rc = 0;
if (tab->at_dovacuum)
rc = snprintf_s(activity, MAX_AUTOVAC_ACTIV_LEN, MAX_AUTOVAC_ACTIV_LEN - 1, "autovacuum: VACUUM%s",
tab->at_doanalyze ? " ANALYZE" : "");
else
rc = snprintf_s(activity, MAX_AUTOVAC_ACTIV_LEN, MAX_AUTOVAC_ACTIV_LEN - 1, "autovacuum: ANALYZE");
securec_check_ss(rc, "\0", "\0");
* Report the qualified name of the relation.
*/
len = strlen(activity);
rc = snprintf_s(activity + len, MAX_AUTOVAC_ACTIV_LEN - len, MAX_AUTOVAC_ACTIV_LEN - len - 1,
" %s.%s%s", tab->at_nspname, tab->at_relname, tab->at_needfreeze ? " (freeze old tuple for recycle clog)" : "");
securec_check_ss(rc, "\0", "\0");
SetCurrentStatementStartTimestamp();
pgstat_report_activity(STATE_RUNNING, activity);
}
* AutoVacuumingActive
* Check GUC vars and report whether the autovacuum process should be
* running.
*/
bool AutoVacuumingActive(void)
{
if (!u_sess->attr.attr_storage.autovacuum_start_daemon || !u_sess->attr.attr_common.pgstat_track_counts ||
SSIsServerModeReadOnly())
return false;
return true;
}
* autovac_init
* This is called at postmaster initialization.
*
* All we do here is annoy the user if he got it wrong.
*/
void autovac_init(void)
{
if (u_sess->attr.attr_storage.autovacuum_start_daemon && !u_sess->attr.attr_common.pgstat_track_counts)
ereport(WARNING, (errmsg("autovacuum not started because of misconfiguration"),
errhint("Enable the \"track_counts\" option.")));
}
* IsAutoVacuum functions
* Return whether this is either a launcher autovacuum process or a worker
* process.
*/
bool IsAutoVacuumLauncherProcess(void)
{
return t_thrd.role == AUTOVACUUM_LAUNCHER;
}
bool IsAutoVacuumWorkerProcess(void)
{
return t_thrd.role == AUTOVACUUM_WORKER;
}
const char* AUTO_VACUUM_WORKER = "AutoVacWorker";
bool IsFromAutoVacWoker(void)
{
return (u_sess->attr.attr_common.application_name != NULL &&
strcmp(u_sess->attr.attr_common.application_name, AUTO_VACUUM_WORKER) == 0);
}
* AutoVacuumShmemSize
* Compute space needed for autovacuum-related shared memory
*/
Size AutoVacuumShmemSize(void)
{
Size size;
* Need the fixed struct and the array of WorkerInfoData.
*/
size = sizeof(AutoVacuumShmemStruct);
size = MAXALIGN(size);
size = add_size(size, mul_size(g_instance.attr.attr_storage.autovacuum_max_workers, sizeof(WorkerInfoData)));
return size;
}
* AutoVacuumShmemInit
* Allocate and initialize autovacuum-related shared memory
*/
void AutoVacuumShmemInit(void)
{
bool found = false;
t_thrd.autovacuum_cxt.AutoVacuumShmem =
(AutoVacuumShmemStruct*)ShmemInitStruct("AutoVacuum Data", AutoVacuumShmemSize(), &found);
if (!IsUnderPostmaster) {
WorkerInfo worker = NULL;
int i = 0;
if (unlikely(found)) {
ereport(PANIC, (errmsg("AutoVacuum Data share mem is already init")));
}
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_launcherpid = 0;
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers = NULL;
SHMQueueInit(&t_thrd.autovacuum_cxt.AutoVacuumShmem->av_runningWorkers);
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_startingWorker = NULL;
worker = (WorkerInfo)((char*)t_thrd.autovacuum_cxt.AutoVacuumShmem + MAXALIGN(sizeof(AutoVacuumShmemStruct)));
for (i = 0; i < g_instance.attr.attr_storage.autovacuum_max_workers; i++) {
worker[i].wi_links.next = (SHM_QUEUE*)t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers;
t_thrd.autovacuum_cxt.AutoVacuumShmem->av_freeWorkers = &worker[i];
}
} else {
if (unlikely(!found)) {
ereport(PANIC, (errmsg("AutoVacuum Data share mem is not init")));
}
}
}
* autovac_refresh_stats
* Refresh pgstats data for an autovacuum process
*
* Cause the next pgstats read operation to obtain fresh data, but throttle
* such refreshing in the autovacuum launcher. This is mostly to avoid
* rereading the pgstats files too many times in quick succession when there
* are many databases.
*
* Note: we avoid throttling in the autovac worker, as it would be
* counterproductive in the recheck logic.
*/
static void autovac_refresh_stats(void)
{
if (IsAutoVacuumLauncherProcess()) {
TimestampTz current_time;
current_time = GetCurrentTimestamp();
if (!TimestampDifferenceExceeds(t_thrd.autovacuum_cxt.last_read, current_time, STATS_READ_DELAY))
return;
t_thrd.autovacuum_cxt.last_read = current_time;
}
pgstat_clear_snapshot();
}
static void partition_needs_vacanalyze(Oid partid, AutoVacOpts* relopts, Form_pg_partition partForm, HeapTuple partTuple,
at_partitioned_table* ap_entry, PgStat_StatTabEntry* tabentry, bool is_recheck, bool* dovacuum,
bool* doanalyze, bool* need_freeze)
{
PgStat_StatTabKey tablekey;
avw_info* avwentry = NULL;
bool found = false;
bool av_enabled = false;
bool force_vacuum = false;
bool delta_vacuum = false;
bytea* partoptions = NULL;
float4 reltuples;
float4 vac_scale_factor = 0.0;
float4 anl_scale_factor = 0.0;
int vac_base_thresh = 0;
int anl_base_thresh = 0;
float4 vacthresh;
float4 anlthresh;
int64 vactuples = 0;
int64 anltuples = 0;
int64 freeze_max_age = 0;
TransactionId xidForceLimit = InvalidTransactionId;
MultiXactId multiForceLimit = InvalidMultiXactId;
char* relname = NULL;
Oid nameSpaceOid = InvalidOid;
char* partname = NULL;
AssertArg(partForm != NULL && OidIsValid(partid));
determine_vacuum_params(vac_scale_factor, vac_base_thresh, anl_scale_factor, anl_base_thresh, freeze_max_age,
av_enabled, xidForceLimit, multiForceLimit, relopts);
if (t_thrd.autovacuum_cxt.recentXid > FirstNormalTransactionId + freeze_max_age)
xidForceLimit = t_thrd.autovacuum_cxt.recentXid - freeze_max_age;
else
xidForceLimit = FirstNormalTransactionId;
bool isNull = false;
TransactionId relfrozenxid = InvalidTransactionId;
Relation rel = heap_open(PartitionRelationId, AccessShareLock);
Datum xid64datum = heap_getattr(partTuple, Anum_pg_partition_relfrozenxid64, RelationGetDescr(rel), &isNull);
if (isNull) {
relfrozenxid = partForm->relfrozenxid;
if (TransactionIdPrecedes(t_thrd.xact_cxt.ShmemVariableCache->nextXid, relfrozenxid) ||
!TransactionIdIsNormal(relfrozenxid)) {
relfrozenxid = FirstNormalTransactionId;
}
} else {
relfrozenxid = DatumGetTransactionId(xid64datum);
}
#ifndef ENABLE_MULTIPLE_NODES
if (!force_vacuum) {
Datum mxidDatum = heap_getattr(partTuple, Anum_pg_partition_relminmxid, RelationGetDescr(rel), &isNull);
MultiXactId relminmxid = isNull ? FirstMultiXactId : DatumGetTransactionId(mxidDatum);
force_vacuum = (MultiXactIdIsValid(relminmxid) && MultiXactIdPrecedes(relminmxid, multiForceLimit));
}
#endif
Datum partoptsdatum = fastgetattr(partTuple, Anum_pg_partition_reloptions, RelationGetDescr(rel), &isNull);
heap_close(rel, AccessShareLock);
force_vacuum = (TransactionIdIsNormal(relfrozenxid) && TransactionIdPrecedes(relfrozenxid, xidForceLimit));
*need_freeze = force_vacuum;
partoptions = heap_reloptions(RELKIND_RELATION, partoptsdatum, false);
if (partoptions != NULL && StdRelOptIsColStore(partoptions) &&
g_instance.attr.attr_storage.enable_delta_store && DO_VACUUM) {
PgStat_StatDBEntry *dbentry = pgstat_fetch_stat_dbentry(u_sess->proc_cxt.MyDatabaseId);;
PgStat_StatDBEntry *shared = pgstat_fetch_stat_dbentry(InvalidOid);
PgStat_StatTabEntry *deltaTabentry = get_pgstat_tabentry_relid(partForm->reldeltarelid, false,
InvalidOid, shared, dbentry);
if (deltaTabentry != NULL) {
delta_vacuum = (deltaTabentry->n_live_tuples >= ((StdRdOptions*)partoptions)->delta_rows_threshold);
}
}
if (!force_vacuum && (!av_enabled || !u_sess->attr.attr_storage.autovacuum_start_daemon)) {
*doanalyze = false;
*dovacuum = false;
return;
}
#ifndef ENABLE_MULTIPLE_NODES
if (!force_vacuum && !(ap_entry->at_allowvacuum && ap_entry->at_dovacuum)) {
#else
if (!force_vacuum && (!ap_entry->at_allowvacuum || ap_entry->at_dovacuum)) {
#endif
*doanalyze = false;
*dovacuum = delta_vacuum;
return;
}
Oid relid = InvalidOid;
if (partForm->parttype == PART_OBJ_TYPE_TABLE_SUB_PARTITION) {
relid = partid_get_parentid(partForm->parentid);
} else {
relid = partForm->parentid;
}
relname = get_rel_name(relid);
nameSpaceOid = get_rel_namespace(relid);
partname = NameStr(partForm->relname);
reltuples = partForm->reltuples;
anlthresh = (float4)anl_base_thresh + anl_scale_factor * reltuples;
vacthresh = (float4)vac_base_thresh + vac_scale_factor * reltuples;
if (NULL != t_thrd.autovacuum_cxt.pgStatAutoVacInfo) {
tablekey.statFlag = relid;
tablekey.tableid = partid;
avwentry =
(avw_info*)hash_search(t_thrd.autovacuum_cxt.pgStatAutoVacInfo, (void*)(&tablekey), HASH_FIND, &found);
}
if ((avwentry == NULL) && (tabentry == NULL)) {
*doanalyze = false;
*dovacuum = force_vacuum || delta_vacuum;
} else {
if (tabentry && (tabentry->changes_since_analyze || tabentry->n_dead_tuples)) {
vactuples = tabentry->n_dead_tuples;
anltuples = tabentry->changes_since_analyze;
AUTOVAC_LOG(DEBUG2, "fetch local stat info: vac \"%s\" partition(\"%s\") "
"changes_since_analyze = %ld n_dead_tuples = %ld ",
relname, partname, tabentry->changes_since_analyze, tabentry->n_dead_tuples);
}
if (avwentry && (avwentry->changes_since_analyze || avwentry->n_dead_tuples)) {
anltuples = avwentry->changes_since_analyze;
vactuples = avwentry->n_dead_tuples;
AUTOVAC_LOG(DEBUG2, "fetch local stat info: vac \"%s\" partition(\"%s\") changes_since_analyze = %ld "
"n_dead_tuples = %ld ", relname, partname, avwentry->changes_since_analyze, avwentry->n_dead_tuples);
* refresh partition's vacthresh/anlthresh with n_live_tuples because we
* do not do analyze for partition, and so reltuples in pg_partition
* is 0, so we just n_live_tuples instead of reltuples
*/
anlthresh = (float4)anl_base_thresh + anl_scale_factor * avwentry->n_live_tuples;
vacthresh = (float4)vac_base_thresh + vac_scale_factor * avwentry->n_live_tuples;
}
*dovacuum = force_vacuum || delta_vacuum;
if (false == *dovacuum)
*dovacuum = (vactuples > vacthresh);
* Determine if this table needs analyze.
* we only do auto-analyze on partitioned table, never on partition
* we keep the code just for we will support partition analyze one day
*/
*doanalyze = (anltuples > anlthresh) && false;
}
if (!is_recheck && (*dovacuum || *doanalyze)) {
AUTOVAC_LOG(DEBUG2, "vac table \"%s\" partition(\"%s\"): recheck = %s need_freeze = %s "
"dovacuum = %s (dead tuples %ld vacuum threshold %.0f)",
relname, partname, is_recheck ? "true" : "false", *need_freeze ? "true" : "false",
*dovacuum ? "true" : "false", vactuples, vacthresh);
}
DEBUG_VACUUM_LOG(relid, nameSpaceOid, LOG, "vac table \"%s\" partition(\"%s\"): recheck = %s "
"need_freeze = %s dovacuum = %s (dead tuples %ld vacuum threshold %.0f) reltuple = %.0f",
relname, partname, is_recheck ? "true" : "false", *need_freeze ? "true" : "false",
*dovacuum ? "true" : "false", vactuples, vacthresh, reltuples);
}
static autovac_table* partition_recheck_autovac(
vacuum_object* vacObj, HTAB* table_relopt_map, HTAB* partitioned_tables_map, TupleDesc pg_class_desc)
{
Oid partid = vacObj->tab_oid;
bool dovacuum = false;
bool doanalyze = false;
bool dovacuum_toast = vacObj->dovacuum_toast;
bool need_freeze = false;
autovac_table* tab = NULL;
PgStat_StatTabEntry* tabentry = NULL;
PgStat_StatDBEntry* shared = NULL;
PgStat_StatDBEntry* dbentry = NULL;
at_partitioned_table* ap_entry = NULL;
AutoVacOpts* avopts = NULL;
av_relation* hentry = NULL;
bool found = false;
Form_pg_partition partForm;
HeapTuple partTuple;
Oid relid;
if (IS_SINGLE_NODE) {
autovac_refresh_stats();
}
partTuple = SearchSysCacheCopy1(PARTRELID, ObjectIdGetDatum(partid));
if (!HeapTupleIsValid(partTuple))
return NULL;
partForm = (Form_pg_partition)GETSTRUCT(partTuple);
if (partForm->parttype == PART_OBJ_TYPE_TABLE_SUB_PARTITION) {
relid = partid_get_parentid(partForm->parentid);
} else {
relid = partForm->parentid;
}
shared = pgstat_fetch_stat_dbentry(InvalidOid);
dbentry = pgstat_fetch_stat_dbentry(u_sess->proc_cxt.MyDatabaseId);
hentry = (av_relation*)hash_search(table_relopt_map, &relid, HASH_FIND, &found);
if (!(found && hentry->ar_hasrelopts)) {
ereport(defence_errlevel(), (errcode(ERRCODE_DATA_CORRUPTED), errmsg("Oid: %u does not "
"find in rel options map.", relid)));
}
avopts = &(hentry->ar_reloptions);
ap_entry = (at_partitioned_table*)hash_search(partitioned_tables_map, &relid, HASH_FIND, &found);
if (!found) {
ereport(defence_errlevel(), (errcode(ERRCODE_DATA_CORRUPTED), errmsg("Oid: %u does not "
"find in partitioned tables map.", relid)));
}
tabentry = get_pgstat_tabentry_relid(partid, false, relid, shared, dbentry);
partition_needs_vacanalyze(
partid, avopts, partForm, partTuple, ap_entry, tabentry, true, &dovacuum, &doanalyze, &need_freeze);
Assert(false == doanalyze);
if (dovacuum || dovacuum_toast) {
tab = calculate_vacuum_cost_and_freezeages(avopts, doanalyze, need_freeze);
if (tab != NULL) {
tab->at_relid = partid;
tab->at_sharedrel = false;
tab->at_dovacuum = dovacuum || dovacuum_toast;
tab->at_gpivacuumed = ap_entry->at_gpivacuumed;
}
}
heap_freetuple(partTuple);
return tab;
}
