[jira] [Commented] (AIRAVATA-2100) Roll out the phase 1 implementation of logging

["Shameera Rathnayaka (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/AIRAVATA-2100?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15538848#comment-15538848 ] 

Shameera Rathnayaka commented on AIRAVATA-2100:
-----------------------------------------------

Hi Lahiru, 

How many instances do you recommend to setup only for log analysis in production with enabling HA? 



> Roll out the phase 1 implementation of logging
> ----------------------------------------------
>
>                 Key: AIRAVATA-2100
>                 URL: https://issues.apache.org/jira/browse/AIRAVATA-2100
>             Project: Airavata
>          Issue Type: Sub-task
>            Reporter: Lahiru Gunathilake
>            Assignee: Lahiru Gunathilake
>
> Following things has to be done to roll out the phase 1 implementation to production.
> 1. Setup a PGA locally or a remote machine.
> 2. Test logging with real experiment runs (not just with startup of the server which I have been doing).
> 3. Once I'm confident commit the changes to develop (Nothing will break, by default kafka logging is disabled)
> 4. Setup a kafka cluster and elastic search cluster for development usage.
> 5. Deploy the develop branch changes with logging enabled (If we already have less critical PGA setup we can use that environment) and test if further before roll out to production. 
> 6. Move the changes to production with a separate elastic search cluster just for production.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)