Fu Wei

ping @ahrtr @serathius @jmhbnz @siyuanfoundation

Hi @ahrtr > Only certain gRPC version (1.60?) has the problem. Current 3.5 and 3.4 are not depending on gRPC 1.60 https://github.com/grpc/grpc-go/pull/6922 fixed gracefulstop issue introduced by v1.60 version. For...

> Because the pipeline is initiated from the inside (node A) to the external nodes (node B and C), thus, it will bypass the proxy, like the red arrow is...

I think we can still follow current design based on L4. However, we need to change `modifyTx` and `modifyRx` functions to support filter packets based on `/proc/net/tcp(6)` (tcp tuple and...

> So , this is not totally network partition, is just some of the flows get partitioned? Hi @aojea yes. Each member has L4 proxy to forward peer's traffic to...

ping @ahrtr @serathius @jmhbnz @siyuanfoundation request a review on this. hope it can help for failpoint test.

+1 for this. Is it good timing to migrate and deprecate all the existing experimental features in v3.6? 👀

@YaoC would you please to send patch to fix this? Thanks!

Thanks @chaochn47 added comment in your reproduce pull request https://github.com/etcd-io/etcd/pull/17535#discussion_r1514119465 > So I would propose prioritize the watch response with CompactRevision is set over the other watch responses so client...

> Do you mean it may cause lost of event? the `syncWatcher` updated the `w.minRev` to the value that is higher than `compactionRev`. The `syncWatcher` doesn't have chance to send...