Todd Lipcon has posted comments on this change. Change subject: KUDU-236. Implement tablet history GC ......................................................................
Patch Set 12: (3 comments) http://gerrit.cloudera.org:8080/#/c/3076/12/src/kudu/integration-tests/tablet_history_gc-itest.cc File src/kudu/integration-tests/tablet_history_gc-itest.cc: Line 30: TEST_F(TabletHistoryGcItest, TestDeleteEmptyTable) { I don't follow the name of this test. A comment explaining its intent would be nice too. Line 33: // Insert some rows. why do we need any rows for this test? http://gerrit.cloudera.org:8080/#/c/3076/12/src/kudu/tablet/compaction.cc File src/kudu/tablet/compaction.cc: Line 1017: // TODO: So are we safe? Actually I'm worried about this... I think if you combine a "missed update" with a GCed row, the row_idxes will get offset and you'll end up re-updating the wrong row. Check out the Test*WithConcurrentMutation tests in tablet-test.cc -- I think we need some kind of coverage of that in coordination with GCed rows. -- To view, visit http://gerrit.cloudera.org:8080/3076 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: If9833a863f118eb82be80ea56204d0d9141611c2 Gerrit-PatchSet: 12 Gerrit-Project: kudu Gerrit-Branch: master Gerrit-Owner: Mike Percy <mpe...@apache.org> Gerrit-Reviewer: Adar Dembo <a...@cloudera.com> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <mpe...@apache.org> Gerrit-Reviewer: Todd Lipcon <t...@apache.org> Gerrit-HasComments: Yes