----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/69076/#review209797 -----------------------------------------------------------
I did a quick look and I feel this logging will be too verbose and may slow the fetch process as there is one message per task executed. There can be thousands or even millions of tasks (one per databse, one per table, one per 100 partitions). Is there another way to print a progress instead of two messages per task? - Sergio Pena On Oct. 18, 2018, 9:29 p.m., Arjun Mishra wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/69076/ > ----------------------------------------------------------- > > (Updated Oct. 18, 2018, 9:29 p.m.) > > > Review request for sentry, kalyan kumar kalvagadda, Na Li, and Sergio Pena. > > > Repository: sentry > > > Description > ------- > > When sentry is fetching snapshot from HMS, it should log periodically on > where it stands in the snapshot process. This will help person debugging it > and help him understand the progress. > > > > This is important as this process could take magnitude of minutes when the > HMS data is huge. > > > Diffs > ----- > > > sentry-service/sentry-service-server/src/main/java/org/apache/sentry/api/service/thrift/SentryMetrics.java > 214d78c53 > > sentry-service/sentry-service-server/src/main/java/org/apache/sentry/service/thrift/FullUpdateInitializer.java > 3e27d1bbe > > sentry-service/sentry-service-server/src/main/java/org/apache/sentry/service/thrift/SentryHMSClient.java > 4baeb6725 > > > Diff: https://reviews.apache.org/r/69076/diff/2/ > > > Testing > ------- > > > Thanks, > > Arjun Mishra > >