errose28 commented on PR #7266:
URL: https://github.com/apache/ozone/pull/7266#issuecomment-2916913677

   Hi @slfan1989 thanks for working on this change. I think there are three 
attributes being added here which should be reviewed separately:
   1. Adding an SCM RPC to retrieve volume information
   2. Tracking failure time of the volume
   3. Adding a CLI to view the volume information
   
   The RPC to retrieve volume information is definitely required going forward 
regardless of the other two items to create some sort of CLI to query volume 
state.
   
   Tracking the failure time of the volume seems like a somewhat invasive 
change since it spans the datanode, heartbeat, and SCM. Is this necessary, or 
is it enough to depend on a metrics database to track timing of cluster events? 
Of course we need improvements to our volume metrics as well as mentioned in 
#8405.
   
   On the CLI front, I do think we need a dedicated `ozone admin datanode info` 
command going forward as outlined in HDDS-13097. This would give all volume 
information per node. With volume counters added to `ozone admin datanode list` 
as proposed in HDDS-13096, we could get all failed volumes in a two step 
process:
   1. jq filter on `ozone admin datanode list` to find all nodes with failed 
volumes.
   2. jq filter on `ozone admin datanode info` to get specific information 
about the failed volumes, including their capacity.
   
   Do we need a dedicated `ozone admin datanode volume list/info` command 
pairing in addition to this? It may be useful to have such cross-cutting 
commands to get information in one shot, but on the other hand it may result in 
duplication at the CLI. For example I could see the request to add node 
filtering to `ozone admin datanode volume list/info` at which point it becomes 
much the same as `ozone admin datanode list/info`.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to