> On Feb. 23, 2016, 6:41 p.m., Alejandro Fernandez wrote: > > Actually, is this because Ranger DB script is called from multiple hosts > > during RU/EU and instead it should only be called once? > > In which case, the fix is to change the orchestration and not the retries.
That is a good suggestion, any lead or any reference to similar implementation? Has any other component taken care of DB operations in HA mode during upgrade? - Gautam ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/43830/#review120359 ----------------------------------------------------------- On Feb. 23, 2016, 12:10 p.m., Gautam Borad wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/43830/ > ----------------------------------------------------------- > > (Updated Feb. 23, 2016, 12:10 p.m.) > > > Review request for Ambari, Alejandro Fernandez, Jaimin Jetly, Jonathan > Hurley, Mahadev Konar, Sumit Mohanty, Selvamohan Neethiraj, and Velmurugan > Periasamy. > > > Bugs: AMBARI-15127 > https://issues.apache.org/jira/browse/AMBARI-15127 > > > Repository: ambari > > > Description > ------- > > Need to apply multiple retries for Ranger DB setup calls to handle concurrent > calls to one DB from 2 different hosts where Ranger Admin is installed. > This scenario is needed to handle upgrades of HDP in a case when Ranger HA is > enabled. > > > Diffs > ----- > > > ambari-server/src/main/resources/common-services/RANGER/0.4.0/package/scripts/setup_ranger_xml.py > 4660cf8 > > Diff: https://reviews.apache.org/r/43830/diff/ > > > Testing > ------- > > Installed Ranger on Debian 7 + Ranger HA and then performed EU to 2.4.0.0-XXX. > > > Thanks, > > Gautam Borad > >
