Alexey Serbin has posted comments on this change. Change subject: Add integration tests for duplicate keys ......................................................................
Patch Set 7: (1 comment) http://gerrit.cloudera.org:8080/#/c/5349/7/src/kudu/integration-tests/cluster_verifier.cc File src/kudu/integration-tests/cluster_verifier.cc: PS7, Line 126: CHECK_OK(scanner.SetReadMode(client::KuduScanner::READ_AT_SNAPSHOT)); : CHECK_OK(scanner.SetFaultTolerant()); : CHECK_OK(scanner.SetTimeoutMillis(5000)); : CHECK_OK(scanner.SetProjectedColumns(vector<string>())); > We were already using this pattern. We RETURN_NOT_OK() on things that may f >From the caller's point of view, what is the difference if this >method/function would return an error either from client->OpenTable() or >scanner.SetReadMode()? I think calling abort() does not make sense when it's possible to handle the error gracefully and staying consistent. As I see it, the only case when you would want to call abort() is when it's clear the consistency of the code is compromised and it's impossible to continue safely. This is not the case, IMO. -- To view, visit http://gerrit.cloudera.org:8080/5349 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I6d018281d412ae034bd7b70c8311077a52b2795d Gerrit-PatchSet: 7 Gerrit-Project: kudu Gerrit-Branch: master Gerrit-Owner: David Ribeiro Alves <dral...@apache.org> Gerrit-Reviewer: Alexey Serbin <aser...@cloudera.com> Gerrit-Reviewer: David Ribeiro Alves <dral...@apache.org> Gerrit-Reviewer: Jean-Daniel Cryans <jdcry...@apache.org> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <mpe...@apache.org> Gerrit-Reviewer: Todd Lipcon <t...@apache.org> Gerrit-HasComments: Yes
