mirror of
https://github.com/kvcache-ai/Mooncake.git
synced 2026-11-05 01:17:55 +00:00
[GH-ISSUE #2210] [Bug]: mooncake master开启etcd高可用模式,etcd下线重连后,mooncake_master系统摘要信息打印错误,之前的client信息重复注册 #607
Labels
No labels
ascend
auto-closed
bug
bug
CI/Build
Common
documentation
enhancement
feature request
good first issue
help wanted
high priority
MetaX
nightly-failure
pull-request
PyTorch Backend
question
RFC
Roadmap
stale
Store
Tests
Transfer Engine
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
kvcache-ai/Mooncake#607
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @1717000160 on GitHub (May 25, 2026).
Original GitHub issue: https://github.com/kvcache-ai/Mooncake/issues/2210
Originally assigned to: @00fish0 on GitHub.
Bug Report
mooncake master开启etcd高可用模式,etcd下线重连后,mooncake_master系统摘要信息打印错误,之前的client信息重复注册
Before submitting...
@1717000160 commented on GitHub (May 25, 2026):
原因是MasterMetricManager是单例模式,MasterService随etcd连接一同创建和销毁,销毁时没有对MasterMetricManager中的信息进行清理,MasterService再次创建是会重复注册client信息
@00fish0 commented on GitHub (May 26, 2026):
感谢反馈!
方便提供一下复现脚本或者复现步骤吗?比如 etcd 下线重连的具体操作方式、master 启动参数、以及触发重复注册时的 client 操作流程。这样我们可以更快定位并验证修复。
@1717000160 commented on GitHub (May 26, 2026):
1.拉起etcd服务: 127.0.0.1:12335
2.拉起master: mooncake_master --eviction_high_watermark_ratio 0.9 --eviction_ratio 0.05 --port 50051 --enable_ha true --etcd_endpoints 127.0.0.1:12335
3.拉起任意一个client注册1GB内存
4.等待master打印中显示client注册成功后,杀掉etcd服务
5.master打印重连日志后,重新拉起etcd
@github-actions[bot] commented on GitHub (Aug 25, 2026):
This issue has had no activity for 90 days and will be closed in 7 days if there is no further activity. Please comment or react if it should stay open.
@github-actions[bot] commented on GitHub (Sep 2, 2026):
Closing due to 3 months of inactivity. If this is still relevant, please comment and we can reopen.