> On Sept. 11, 2015, 2:51 p.m., Jonathan Hurley wrote:
> > Why is this change needed for stop-the-world? You've already stopped the 
> > hive server and prevented clients from draining. You should not need to 
> > adjust the ports like we do for normal rolling upgrades.

You're right. Stop-the-World Upgrade doesn't need to temporarily change the 
hive port.


- Alejandro


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/38220/#review98609
-----------------------------------------------------------


On Sept. 9, 2015, 2:52 p.m., Dmitro Lisnichenko wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/38220/
> -----------------------------------------------------------
> 
> (Updated Sept. 9, 2015, 2:52 p.m.)
> 
> 
> Review request for Ambari, Alejandro Fernandez, Dmytro Grinenko, Jonathan 
> Hurley, and Nate Cole.
> 
> 
> Bugs: AMBARI-13048
>     https://issues.apache.org/jira/browse/AMBARI-13048
> 
> 
> Repository: ambari
> 
> 
> Description
> -------
> 
> Goal:
> for STW Upgrade, we'll need to apply configs after all services have been 
> stopped, and before calling hdp-select set all.
> 
> I think, we have to add separate groups before and after restart of HIVE 
> server to enforce config application order during UPGRADE and DONGRADE. Looks 
> like that is the only config change required for 2.2->2.2+ upgrade. Does such 
> upgrade pack definition look good?
> 
> 
> Diffs
> -----
> 
>   
> ambari-server/src/main/resources/stacks/HDP/2.2/upgrades/nonrolling-upgrade-2.2.xml
>  01022b8 
> 
> Diff: https://reviews.apache.org/r/38220/diff/
> 
> 
> Testing
> -------
> 
> Preview
> 
> 
> Thanks,
> 
> Dmitro Lisnichenko
> 
>

Reply via email to