| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Lockfree connection pool (#54) * Modify connection pool to lockfree. * Use blocking queue in connection. * Dequeue limited number of reqs to alleviate hol blocking. --------- Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 8 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Lockfree connection pool (#54) * Modify connection pool to lockfree. * Use blocking queue in connection. * Dequeue limited number of reqs to alleviate hol blocking. --------- Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 8 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 | |
Merge dev into main (#37) * Feature monitor (#19) * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) * Implement cli stats and stats-all. Implement prometheus monitor. * Modify prometheus monitor to report avg latency. Add ut for stats. * Fix bug of displaying stats. * Fix bug in display. * Fix bug of wrong scatter in stats and stats-all. * Fix wait time in prometheus. * Modify stat-max to default false. * Add compile option for prometheus. * Add prometheus to docker file. * Add prometheus build to docker build script. * Add prometheus dependency to all docker files. Add license. * Modify default config. --------- Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * When invoking cache eviction, quickly ignore the item which cannot be evicted (#24) (#28) Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> * add meta error report (#25) * add the error report for meta server * enable error report through env * add config for meta server replication error report * Adapt car docker image. (#33) Adapt car prometheus report. Set max open instance num in config file. Add log in client side meta error. Change to timed wait. Report current open fds. Set max open instance to 0 in config will unset the limit. Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> * Add libfs test framework. (#36) --------- Co-authored-by: qiuziyan1998 <62130191+qiuziyan1998@users.noreply.github.com> Co-authored-by: Lu Zhang <lu_zhang92@outlook.com> Co-authored-by: qiuziyan1998 <qiuziyan1@huawei.com> | 11 个月前 |
Remote test
Set ALL_SERVERS to actual remote test nodes in distribute.sh, kill_test.sh, remote_run.sh. Enable ssh connection among all nodes, and append the public key of the client node running remote_run.sh to authorized_keys of the nodes in ALL_SERVERS. Modify parameters in remote_run.sh to actual values in used.
bash remote-run.sh
The IOPS throughput will be printed.
Local test
Modify parameters in local_run.sh to actual values in used.
bash local-run.sh
The IOPS throughput will be printed.