> On Sept. 6, 2017, 2:02 a.m., Vamsee Yarlagadda wrote: > > sentry-hdfs/sentry-hdfs-service/src/main/java/org/apache/sentry/hdfs/DBUpdateForwarder.java > > Line 115 (original), 122 (patched) > > <https://reviews.apache.org/r/62107/diff/1/?file=1816055#file1816055line127> > > > > What would prevent the execution from reaching this step and a full > > snapshot from HMS kicked off at the same time? > > > > Do you think we should have the block on both the sides? > > Brian Towles wrote: > Once it hits this point all of the other cases have been tried and the > full update should be available due to an unavailable sequence number
Agreed but my question is let's say HDFS request (after going through all the condition checks) decides to send a full snapshot and there could be some activity on HMSFollower side that can also start a full snapshot from HMS. There is nothing that is stopping HMSFollower from doing that. During this time, the memory consumption will spike again because of both the snapshots happening at the same time. - Vamsee ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/62107/#review184619 ----------------------------------------------------------- On Sept. 6, 2017, 3:52 p.m., Brian Towles wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/62107/ > ----------------------------------------------------------- > > (Updated Sept. 6, 2017, 3:52 p.m.) > > > Review request for sentry, Alexander Kolbasov, Arjun Mishra, Na Li, Sergio > Pena, and Vadim Spector. > > > Bugs: SENTRY-1918 > https://issues.apache.org/jira/browse/SENTRY-1918 > > > Repository: sentry > > > Description > ------- > > SENTRY-1918: NN snapshot should not be served while HMS snapshot is collected > > Added fix to limit sending back of full snapshot when the HMSFollower is > pulling a full snapshot > > > Diffs > ----- > > > sentry-hdfs/sentry-hdfs-service/src/main/java/org/apache/sentry/hdfs/DBUpdateForwarder.java > 5d744214e14d6c48194b3a0bf84d6e10070b020a > > sentry-hdfs/sentry-hdfs-service/src/test/java/org/apache/sentry/hdfs/TestDBUpdateForwarder.java > 8fbc10048003ab4b8a38676e241ae0fd27d2392c > > sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java > d4feb380fa0f3bd5f237609a107295a2d23eae3b > > sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/SentryService.java > d44abffc199cd46d3ab7d2230dfdb2075736a8f0 > > sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/SentryServiceFactory.java > 1685702c6dbc9715c8885a29a80bc68509550f0b > > > Diff: https://reviews.apache.org/r/62107/diff/4/ > > > Testing > ------- > > Unit Tests > > > Thanks, > > Brian Towles > >
