[ 
https://issues.apache.org/jira/browse/AMBARI-19046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Thomas Larsson updated AMBARI-19046:
------------------------------------
    Description: 
When enabling ResourceManager HA, alerts such as "ResourceManager CPU 
utilisation" become aware that there are two RM's to check.

However, when the alert queries the standby RM it will just respond with a http 
307 redirect and the alert will end up in an "Unknown" state.

This means that after RM HA is enabled, either you live with these alert 
constantly alerting, or you disable them.

Perhaps this is unsolvable without changes to the RM but I'll create the ticket 
here as the first place.

The alerts that we had to disable are:

* ResourceManager CPU Utilization
* ResourceManager RPC Latency
* NodeManager Health Summary


  was:
When enabling ResourceManager HA, alerts such as "ResourceManager CPU 
utilisation" become aware that there are two RM's to check.

However, when the alert queries the standby RM it will just respond with a http 
307 redirect and the alert will end up in an "Unknown" state.

This means that after RM HA is enabled, either you live with these alert 
constantly alerting, or you disable them.

Perhaps this is unsolvable without changes to the RM but I'll create the ticket 
here as the first place.


> Ambari ResourceManager alerts not working with ResourceManager HA
> -----------------------------------------------------------------
>
>                 Key: AMBARI-19046
>                 URL: https://issues.apache.org/jira/browse/AMBARI-19046
>             Project: Ambari
>          Issue Type: Bug
>          Components: alerts
>    Affects Versions: 2.2.2
>            Reporter: Thomas Larsson
>            Priority: Minor
>
> When enabling ResourceManager HA, alerts such as "ResourceManager CPU 
> utilisation" become aware that there are two RM's to check.
> However, when the alert queries the standby RM it will just respond with a 
> http 307 redirect and the alert will end up in an "Unknown" state.
> This means that after RM HA is enabled, either you live with these alert 
> constantly alerting, or you disable them.
> Perhaps this is unsolvable without changes to the RM but I'll create the 
> ticket here as the first place.
> The alerts that we had to disable are:
> * ResourceManager CPU Utilization
> * ResourceManager RPC Latency
> * NodeManager Health Summary



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

Reply via email to