Alexey Serbin has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/18029 )

Change subject: [tools] KUDU-3337 Add unsafe_create_cmeta tool
......................................................................


Patch Set 6:

(3 comments)

http://gerrit.cloudera.org:8080/#/c/18029/6/src/kudu/tools/kudu-tool-test.cc
File src/kudu/tools/kudu-tool-test.cc:

http://gerrit.cloudera.org:8080/#/c/18029/6/src/kudu/tools/kudu-tool-test.cc@18
PS6, Line 18: #include <limits.h>
nit: replace with <climits> C++ header file


http://gerrit.cloudera.org:8080/#/c/18029/3/src/kudu/tools/tool_action_local_replica.cc
File src/kudu/tools/tool_action_local_replica.cc:

http://gerrit.cloudera.org:8080/#/c/18029/3/src/kudu/tools/tool_action_local_replica.cc@675
PS3, Line 675:
> If the tablet in question was a non-voter, it wouldn't affect consensus, an
I guess it's possible to have a situation when there was still a majority of 
replicas when a non-voter replica had been added and tablet copying started, 
but during the catch-up phase the majority wasn't present anymore.  This 
situation could persist for some time: such a non-voter replica would be 
evicted, but only if there is a majority to update Raft config.

Even such a situation is of rare occurrence, I was curious whether we target 
this tool to help in such cases as well.  Say, the only alive replicas out of 
three are: (1) a voter (2) a non-voter stuck in such limbo.  And this 
particular former replica was a voter, but now its Raft metadata file is 
corrupted, and there is an attempt to fix the issue.

It's totally fine with me if such a situation is out of scope for this patch, 
and it might be a follow-up one to address such a case separately.


http://gerrit.cloudera.org:8080/#/c/18029/3/src/kudu/tools/tool_action_local_replica.cc@678
PS3, Line 678: Status UnsafeRecreateCmeta(const RunnerContext& context) {
             :   const string& tablet_id = FindOrDie(context.required_args, 
kTabletIdArg);
             :   int64_t term;
             :   int64_t opid_index;
             :   try {
> Does it make sense to add an extra check for the arguments provided w.r.t.
It seems this comment isn't addressed as of PS6.  Or the idea is to address it 
in a follow-up PS or even in a separate changelist?



--
To view, visit http://gerrit.cloudera.org:8080/18029
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: kudu
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I136cc5b5797420a9ca9156f37c3e281da0c265d7
Gerrit-Change-Number: 18029
Gerrit-PatchSet: 6
Gerrit-Owner: Attila Bukor <[email protected]>
Gerrit-Reviewer: Abhishek Chennaka <[email protected]>
Gerrit-Reviewer: Alexey Serbin <[email protected]>
Gerrit-Reviewer: Andrew Wong <[email protected]>
Gerrit-Reviewer: Attila Bukor <[email protected]>
Gerrit-Reviewer: Kudu Jenkins (120)
Gerrit-Reviewer: Zoltan Chovan <[email protected]>
Gerrit-Comment-Date: Mon, 06 Nov 2023 20:50:46 +0000
Gerrit-HasComments: Yes

Reply via email to