> On Jan. 11, 2017, 9:24 p.m., Alejandro Fernandez wrote: > > ambari-server/src/main/resources/common-services/ATLAS/0.1.0.2.3/package/scripts/service_check.py, > > line 37 > > <https://reviews.apache.org/r/55382/diff/1/?file=1601257#file1601257line37> > > > > We need to be able to tolerate a max of 10-20% down nodes. > > Service Checks need to be robust to these type of intermittent failures. > > If just 1/300 hosts Atlas hosts is bad, I don't want this to cause > > failures for features like Rolling Upgrade.
@Alejandro - what should the service_check do? Break the loop after first success? - Madhan ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/55382/#review161297 ----------------------------------------------------------- On Jan. 10, 2017, 3:54 p.m., Vishal Suvagia wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/55382/ > ----------------------------------------------------------- > > (Updated Jan. 10, 2017, 3:54 p.m.) > > > Review request for Ambari, Alejandro Fernandez, Gautam Borad, Jaimin Jetly, > Madhan Neethiraj, Mugdha Varadkar, and Oliver Szabo. > > > Bugs: AMBARI-19386 > https://issues.apache.org/jira/browse/AMBARI-19386 > > > Repository: ambari > > > Description > ------- > > Atlas service check fails as its hard coded to run check against the same > host where service check is running. It should instead run check against > ATLAS metadata server host. > > > Diffs > ----- > > > ambari-server/src/main/resources/common-services/ATLAS/0.1.0.2.3/package/scripts/params.py > 6a53b1a > > ambari-server/src/main/resources/common-services/ATLAS/0.1.0.2.3/package/scripts/service_check.py > 8e27b38 > > Diff: https://reviews.apache.org/r/55382/diff/ > > > Testing > ------- > > Verified install on centos6. > > > Thanks, > > Vishal Suvagia > >