Wei-Chiu Chuang created HDDS-14501:
--------------------------------------
Summary: [Website v2] [Docs] [Administrator Guide] Replacing
Datanode Disks
Key: HDDS-14501
URL: https://issues.apache.org/jira/browse/HDDS-14501
Project: Apache Ozone
Issue Type: Sub-task
Components: documentation
Reporter: Wei-Chiu Chuang
[https://ozone-site-v2.staged.apache.org/docs/administrator-guide/operations/disk-replacement/datanodes]
A datanode may have multiple data volumes, specified in hdds.datanode.dir. For
example,
{noformat}
/data1,/data2,/data3{noformat}
hdds.datanode.failed.data.volumes.tolerated: The number of data volumes that
are allowed to fail before a datanode stops offering service. By default, this
value is -1, meaning unlimited.
Similarly, hdds.datanode.failed.metadata.volumes.tolerated allows a number of
metadata volumes to fail.
During datanode startup, it performs check to determine if a volume fails. If
the datanode is allowed to continue without abort, the volume is taken off.
After datanode starts, a periodic disk check is run every 60 minutes
(determined by configuration property
hdds.datanode.periodic.disk.check.interval.minutes.
When a volume is determined failed, it is chosen by volume choosing policy to
allocate new containers.
To replace the failed disks, shut down the datanode, update hdds.datanode.dir
to remove it from the directory list, and then restart the datanode.
note: Ozone datanode does not support hotswap yet, meaning to update the disk
list, it must restart the datanode process.
The state of volumes can be seen in Datanode metrics and web UI.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
