Pull Request has been successfully merged by CANN-robot.
（Thank you for 黄建波's contribution.）
变更摘要
本 PR 旨在修复算子重试（Op Retry）场景下的问题，改动集中在 src/legacy/ascend910/framework/cluster_maintenance/recovery/operator_retry/opretry_server.cc 中 OpRetryServerHandleError::ProcessEvent 方法的错误日志。原日志笼统地提示"超时"（OpRetryServerHandleError timeout），现改为更明确地说明"aicpu 无法重试，opname 不一致"，使该分支的失败原因（操作名不一致导致无法重试）在日志中清晰可辨，便于问题定位。
主要改动
- 错误日志信息修正：在
OpRetryServerHandleError::ProcessEvent的超时分支中，将日志文案由OpRetryServerHandleError timeout替换为aicpu can not retry, opname is inconsistent，准确反映该分支的真实失败原因（opname 不一致导致 aicpu 无法重试）。 - 日志可观测性提升：新日志使运维人员无需排查上下文即可直接从错误信息判断为算子名不一致导致的重试失败，降低定位成本。
Thanks for your pull-request.
The full list of commands accepted by me can be found at here.
You can get sig-info at here.
You can self-configure the PR merge rules for this repository. For more details, please refer to Here.
For more, you also can visit HICANN.
PR Approval Progress
✅ Congratulations! All modules have met the lgtm and approve requirements.
Module Approval Details
| module | lgtm status | approve status |
|---|---|---|
| src/legacy/ascend910/framework | ✅ 陈浩, 杨邵华 (2/2) | ✅ 陈浩 (1/1) |
💡 Tip:
- Committer can comment
/approveor/lgtm- Commenting
/approveimplies both code review (lgtm) and intent to merge (approve)
CLA Signature Pass
muzhiyu6, thanks for your pull request. All authors of the commits have signed the CLA. 👍
CANN-robot
compile
/lgtm
/approve
描述
batch_send_recv算子注入故障之后没有约束打印，aicpu侧等待stop命令等待超时报错。
修改方案: batch_send_recv 故障算子index不一致，无法重执行，增加约束打印
变更类型
请选择本次引入的变更类型：
关联的Issue
https://gitcode.com/cann/hcomm/issues/814
测试
已完成的测试用例和场景：
补充的UT用例：
文档更新
合入检查