Adar Dembo has posted comments on this change. ( http://gerrit.cloudera.org:8080/9393 )
Change subject: KUDU-2290: Tool to re-create a tablet ...................................................................... Patch Set 8: (2 comments) http://gerrit.cloudera.org:8080/#/c/9393/7/src/kudu/integration-tests/CMakeLists.txt File src/kudu/integration-tests/CMakeLists.txt: http://gerrit.cloudera.org:8080/#/c/9393/7/src/kudu/integration-tests/CMakeLists.txt@101 PS7, Line 101: ADD_KUDU_TEST(replace_tablet-itest) > B-but the test always fails and is disabled, because of KUDU-2376. Can I ad Sorry, I missed that the disabled test was the only test in the file. Add the TODO near the test itself so it'll be more obvious when it is reenabled. http://gerrit.cloudera.org:8080/#/c/9393/7/src/kudu/master/catalog_manager.cc File src/kudu/master/catalog_manager.cc: http://gerrit.cloudera.org:8080/#/c/9393/7/src/kudu/master/catalog_manager.cc@4376 PS7, Line 4376: LOG(INFO) << "ReplaceTablet: tablet " << old_tablet->id() : << " deleted and replaced by tablet " << new_tablet->id(); : resp->set_replaceme > Does the following order work? 1. l_table doesn't actually need to be in WRITE mode; READ is sufficient. That should simplify this a bit by eliminating step #4. 2. We should be able to follow what ProcessPendingAssignments does, right? It also replaces tablets. Looks like it first commits the tablets lock, then adds/removes the tablets to the table, then finally updates the tablet map. In general I'm not sure there's a deadlock free way to do this without exposing a window where GetTableLocations and GetTabletLocations aren't totally consistent with one another. From the looks of it, ProcessPendingAssignments is vulnerable to the window you're describing. Maybe it's able to get away with it because the table is expected to be generally unusable during this time (it's just after a CreateTable after all). Though, maybe the same code path is followed during an ALTER TABLE ADD PARTITION too? -- To view, visit http://gerrit.cloudera.org:8080/9393 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: kudu Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Ifbefbde68e3ca724f04efe0426a3906e5c33ed3c Gerrit-Change-Number: 9393 Gerrit-PatchSet: 8 Gerrit-Owner: Will Berkeley <wdberke...@gmail.com> Gerrit-Reviewer: Adar Dembo <a...@cloudera.com> Gerrit-Reviewer: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: Dan Burkert <danburk...@apache.org> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Tidy Bot Gerrit-Reviewer: Todd Lipcon <t...@apache.org> Gerrit-Reviewer: Will Berkeley <wdberke...@gmail.com> Gerrit-Comment-Date: Tue, 03 Apr 2018 04:09:11 +0000 Gerrit-HasComments: Yes
