[
https://issues.apache.org/jira/browse/HDDS-3831?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17142014#comment-17142014
]
Stephen O'Donnell commented on HDDS-3831:
-----------------------------------------
Currently the container endpoint has:
getContainers - list all containers in batches (/containers)
getKeys - List all keys in a container (/containers/{id}/keys)
getReplicaHistory - History of locations (/containers/{id}/replicaHistory)
getMissing - List of missing containers (/containers/missing)
Now that we have different types of unhealthy containers it seems we need some
more flexibility around reporting on them, but it also depends on what we want
to display in the recon API.
It feels like we should have a summary report for unhealthy containers, eg:
missing: x
mis-replicated: x
under-replicated: x
over-replicated: x
Or perhaps an overall report (this could be part of the getContainers or
another new API):
totalContainers: xxx
healthy: xxx
missing: xxx
mis-replicated: xxx
under-replicated: xxx
over-replicated: xxx
We also probably want to be able to list the unhealthy containers in a given
state, and it probably should be a batch API to limit the number returned if
there are a lot of them.
I wonder if we should have a new endpoint which is:
/containers/unhealthy/{state}
By default it will return a list of all unhealthy containers
(/containers/unhealthy), but this can be filtered by adding MISSING /
UNDER-REPLICATED, OVER-REPLICATED, MIS-REPLICATED.
Then we can remove the existing "missing" endpoint.
[~avijayan] [~vivekratnavel] - what do you guys think, keeping what we will
want to display on Recon in mind?
> Enhance Recon ContainerEndPoint to report on difference unhealty container
> states
> ---------------------------------------------------------------------------------
>
> Key: HDDS-3831
> URL: https://issues.apache.org/jira/browse/HDDS-3831
> Project: Hadoop Distributed Data Store
> Issue Type: Improvement
> Components: Ozone Recon
> Affects Versions: 0.6.0
> Reporter: Stephen O'Donnell
> Priority: Major
>
> HDDS-3082 changed the unhealthyContainers job to store the counts of under
> replicated, over replicated, missing and mis-replicated containers.
> We should enhance ContainerEndPoint.java to provide API access to those
> states and perhaps a summary status too, eg:
> under-replicated-count: 10
> over-replication-count: 0
> mis-replicated-count: 5
> ...
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: ozone-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: ozone-issues-h...@hadoop.apache.org
