Andrew Wong has posted comments on this change. ( http://gerrit.cloudera.org:8080/15323 )
Change subject: ksck: display quiecing-related info ...................................................................... Patch Set 2: (6 comments) http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/integration-tests/tablet_server_quiescing-itest.cc File src/kudu/integration-tests/tablet_server_quiescing-itest.cc: http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/integration-tests/tablet_server_quiescing-itest.cc@398 PS1, Line 398: "-----------+----------------+-----------------\n" : " true | 1 | 0"); : ASSERT_TRUE(ts->server()->quiescing()); : : // Same with ksck. : ASSERT_OK(RunKuduTool({ "cluster", "ksck", master_addr }, &stdout)); : ASSERT_STR_MATCHES(stdout, : ".* Quiescing | Tablet Leaders | Active Scanners\n" : ".*-----------+----------------+-----------------\n" : ".* true | 1 | 0"); : ASSERT_TRUE(ts->server()->quiescing()); > nit: maybe, make the names of the corresponding columns match for both 'tse Done http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/integration-tests/tablet_server_quiescing-itest.cc@433 PS1, Line 433: ; > Instead, maybe output the quiescing information only if --quiescing_info fl I'm a bit conflicted about this. It's not information that you'd really think to look for without extra context about a rolling restart, but I think it'll be valuable for cases even beyond just debugging a rolling restart. It makes `tserver quiesce status` less valuable for sure, but I don't think it makes it entirely useless since running it without the rest of `ksck` might be preferred (especially on larger clusters). I'll keep this as is for now, curious whether you think this _isn't_ worth having by default, or whether this comment is more towards reducing duplication of `tserver quiesce status`. http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/tools/ksck_remote.cc File src/kudu/tools/ksck_remote.cc: http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/tools/ksck_remote.cc@81 PS1, Line 81: true > Do we really want to include the information into ksck by default? Given t Re: backwards compatibility, I don't think it's unreasonable to just log a message noting the lack of support. I hope the more common case will be that it'll be run against newer versions of Kudu. Re: the sub-command, I mentioned this on your other comment, but I think it's still valuable to have that tool as separate from ksck, since ksck is a pretty heavy-weight operation. That said, it is our go-to when it comes to understanding what's going on in a cluster, and quiescing info is a huge value add IMO when it comes to understanding performance and workload skew. http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/tools/ksck_remote.cc@82 PS1, Line 82: to displ > What does 'check' means here? Simply 'display'? Done http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/tools/ksck_remote.cc@298 PS1, Line 298: quiescing_info_ = qinfo; > warning: std::move of the variable 'qinfo' of the trivially-copyable type ' Done http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/tserver/tablet_service.cc File src/kudu/tserver/tablet_service.cc: http://gerrit.cloudera.org:8080/#/c/15323/1/src/kudu/tserver/tablet_service.cc@1020 PS1, Line 1020: case TabletServerFeatures::QUIESCING: > What about other already supported features? Done -- To view, visit http://gerrit.cloudera.org:8080/15323 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: kudu Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Ibdc650eb3ee30e8993330f2cbd389076ea2bad49 Gerrit-Change-Number: 15323 Gerrit-PatchSet: 2 Gerrit-Owner: Andrew Wong <[email protected]> Gerrit-Reviewer: Adar Dembo <[email protected]> Gerrit-Reviewer: Alexey Serbin <[email protected]> Gerrit-Reviewer: Andrew Wong <[email protected]> Gerrit-Reviewer: Kudu Jenkins (120) Gerrit-Reviewer: Tidy Bot (241) Gerrit-Comment-Date: Mon, 02 Mar 2020 22:23:19 +0000 Gerrit-HasComments: Yes
