Adar Dembo has posted comments on this change.

Change subject: Refactor retry handling logic for writes
......................................................................


Patch Set 7:

(3 comments)

http://gerrit.cloudera.org:8080/#/c/2970/7/src/kudu/client/batcher.cc
File src/kudu/client/batcher.cc:

Line 213: can be used to choose the action to take
> well it doesn't tell the caller to empty the cache or retry, just puts the 
OK, fair enough.


http://gerrit.cloudera.org:8080/#/c/2970/7/src/kudu/rpc/replicated_rpc.h
File src/kudu/rpc/replicated_rpc.h:

Line 22: // Result status of a replicated Rpc.
> I called them replicated because they target a replicated resource (tablet,
OK, but scans also target replicas, and yet ReplicatedRpcStatus doesn't apply 
to them. So I think we need to be more precise in our terminology here so it's 
clear why this applies to writes but not scans.


Line 23: struct ReplicatedRpcStatus {
> Not sure, likely not. I see us using this for "writes" in the broad sense (
I won't lie; I was hoping you'd take the refactoring further and find a good 
way to unify analyze/retry between scans and writes. But I guess we shouldn't 
let the perfect be the enemy of the good...


-- 
To view, visit http://gerrit.cloudera.org:8080/2970
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I0e0d491f902191c88c58e3d627106cc1be1bb3cc
Gerrit-PatchSet: 7
Gerrit-Project: kudu
Gerrit-Branch: master
Gerrit-Owner: David Ribeiro Alves <[email protected]>
Gerrit-Reviewer: Adar Dembo <[email protected]>
Gerrit-Reviewer: David Ribeiro Alves <[email protected]>
Gerrit-Reviewer: Kudu Jenkins
Gerrit-Reviewer: Todd Lipcon <[email protected]>
Gerrit-HasComments: Yes

Reply via email to