Updated servePeers to remove the grpc server

[ahrtr]

Fix issues/13564

@gyuho @serathius @ptabor

[ahrtr]

Just rebased the PR.

[ptabor]

That's pretty big change.

• I digged some context for this: corruption alarm #8420, seems to be added for hashkv support.

I agree that wide access to this port is misleading and promotes bad usage patterns.

If we want to proceed I think we should mention it as an explicit breaking change in the Changelog as some user's might depend on this behavior.

[ahrtr]

Thanks @ptabor for the feedback. Just added a breaking change item.

[ahrtr]

@ptabor I removed the gRPC server from peer listener in the integration test as well. All tests passed. Usually users do not know that the 2380 port can also serve client requests because there is no any documentation on this. So I think it might be safe to remove the misleading functionality.

cc @serathius @spzala

[serathius]

LGTM

[ptabor]

It seems it could break --experimental-corrupt-check-time. Maybe this mode is not properly tested.

[ahrtr]

It seems it could break --experimental-corrupt-check-time. Maybe this mode is not properly tested.

@ptabor I tested manually in my local environment, and confirmed that it doesn't break the corruption check functionality.

Firstly I build the etcd binary using my branch ahrtr:remove_peer_serve_client_requests.

Secondly, start a etcd cluster with only 2 members (note that I already got the etcd/bin directory included in the $PATH),

etcd --name infra1 --listen-client-urls http://127.0.0.1:2379 --advertise-client-urls http://127.0.0.1:2379 --listen-peer-urls http://127.0.0.1:12380 --initial-advertise-peer-urls http://127.0.0.1:12380 --initial-cluster-token etcd-cluster-1 --initial-cluster 'infra1=http://127.0.0.1:12380,infra2=http://127.0.0.1:22380' --initial-cluster-state new --enable-pprof --logger=zap --log-outputs=stderr --experimental-corrupt-check-time 5s

etcd --name infra2 --listen-client-urls http://127.0.0.1:22379 --advertise-client-urls http://127.0.0.1:22379 --listen-peer-urls http://127.0.0.1:22380 --initial-advertise-peer-urls http://127.0.0.1:22380 --initial-cluster-token etcd-cluster-1 --initial-cluster 'infra1=http://127.0.0.1:12380,infra2=http://127.0.0.1:22380' --initial-cluster-state new --enable-pprof --logger=zap --log-outputs=stderr --experimental-corrupt-check-time 5s

Thirdly, confirmed that infra1 is the leader,

$ etcdctl  --endpoints http://127.0.0.1:2379,http://127.0.0.1:22379 endpoint status -w table
+------------------------+------------------+-----------+---------+-----------+------------+-----------+------------+--------------------+--------+
|        ENDPOINT        |        ID        |  VERSION  | DB SIZE | IS LEADER | IS LEARNER | RAFT TERM | RAFT INDEX | RAFT APPLIED INDEX | ERRORS |
+------------------------+------------------+-----------+---------+-----------+------------+-----------+------------+--------------------+--------+
|  http://127.0.0.1:2379 | 8211f1d0f64f3269 | 3.6.0-pre |   25 kB |      true |      false |         2 |          7 |                  7 |        |
| http://127.0.0.1:22379 | 91bc3c398fb3c146 | 3.6.0-pre |   25 kB |     false |      false |         2 |          7 |                  7 |        |
+------------------------+------------------+-----------+---------+-----------+------------+-----------+------------+--------------------+--------+

Lastly, the leader prints the following log entries periodically (every 5 seconds),

{"level":"info","ts":"2022-01-26T15:24:11.112+0800","caller":"etcdserver/corrupt.go:244","msg":"finished peer corruption check","number-of-peers-checked":1}
{"level":"info","ts":"2022-01-26T15:24:16.115+0800","caller":"etcdserver/corrupt.go:244","msg":"finished peer corruption check","number-of-peers-checked":1}
{"level":"info","ts":"2022-01-26T15:24:21.119+0800","caller":"etcdserver/corrupt.go:244","msg":"finished peer corruption check","number-of-peers-checked":1}
{"level":"info","ts":"2022-01-26T15:24:26.124+0800","caller":"etcdserver/corrupt.go:244","msg":"finished peer corruption check","number-of-peers-checked":1}
{"level":"info","ts":"2022-01-26T15:24:31.126+0800","caller":"etcdserver/corrupt.go:244","msg":"finished peer corruption check","number-of-peers-checked":1}
......

BTW, I rebased the PR. Please let me know if you have other concerns or comments.

[ahrtr]

cc @spzala

[ahrtr]

Just rebased this PR.

[ahrtr]

FYI, the command etcdctl endpoint hashkv --cluster is also working well on this PR, @ptabor

$ etcdctl endpoint hashkv --cluster -w table
+------------------------+------------+
|        ENDPOINT        |    HASH    |
+------------------------+------------+
|  http://127.0.0.1:2379 | 1084519789 |
| http://127.0.0.1:22379 | 1084519789 |
| http://127.0.0.1:32379 | 1084519789 |
+------------------------+------------+

[ahrtr]

Hi @ptabor , Do you still have any concern? cc @serathius @spzala

[serathius]

@ptabor any other concerns?

[ptabor]

Thank you for checking.