[
https://issues.apache.org/jira/browse/HDDS-16663?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ivan Andika updated HDDS-16663:
-------------------------------
Description:
This is similar to HDFS-5153 to handle large number of containers per volume.
Since disk manufacturers kept pushing the capacity per volume (20TB is now
quite normal with some disk manufacturers to push 40TB and even more) by using
techinques like SAMR, HAMR or increasing the number of disk platters, the
number of containers per volume can increase quite a bit. In a datanodes with a
high number of volumes, we can encounter the same issues in HDFS-5153.
Currently our cluster has not encountered the bottlenecks so this can be
addressed in the future.
was:
This is similar to HDFS-5153 to handle large number of containers per volume.
Since disk manufacturers kept pushing the capacity per volume (20TB is now
quite normal with some disk manufacturers to push 40TB and even more), the
number of containers per volume can increase quite a bit. In a datanodes with a
high number of volumes, we can encounter the same issues in HDFS-5153.
Currently our cluster has not encountered the bottlenecks so this can be
addressed in the future.
> Datanode should send block reports for each storage in a separate message
> -------------------------------------------------------------------------
>
> Key: HDDS-16663
> URL: https://issues.apache.org/jira/browse/HDDS-16663
> Project: Apache Ozone
> Issue Type: Sub-task
> Reporter: Ivan Andika
> Assignee: Ivan Andika
> Priority: Major
>
> This is similar to HDFS-5153 to handle large number of containers per volume.
> Since disk manufacturers kept pushing the capacity per volume (20TB is now
> quite normal with some disk manufacturers to push 40TB and even more) by
> using techinques like SAMR, HAMR or increasing the number of disk platters,
> the number of containers per volume can increase quite a bit. In a datanodes
> with a high number of volumes, we can encounter the same issues in HDFS-5153.
> Currently our cluster has not encountered the bottlenecks so this can be
> addressed in the future.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
