[
https://issues.apache.org/jira/browse/HDFS-15900?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17303885#comment-17303885
]
Akira Ajisaka commented on HDFS-15900:
--------------------------------------
bq. if it's ok to have several records sharing the same nameserviceId in
activeNamespaces,
IMO, there may be multiple active NameNodes if RBF supports Observer NameNodes
in the future, so it's okay to have several records sharing the same
nameserviceId in active Namespaces. However, it's not okay to have UNAVAILABLE
NameNodes registrations in activeNamespaces (I used "we expect" because the
source code is written in that way, sorry for the confusion).
> RBF: empty blockpool id on dfsrouter caused by UNAVAILABLE NameNode
> -------------------------------------------------------------------
>
> Key: HDFS-15900
> URL: https://issues.apache.org/jira/browse/HDFS-15900
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: rbf
> Affects Versions: 3.3.0
> Reporter: Harunobu Daikoku
> Assignee: Harunobu Daikoku
> Priority: Major
> Attachments: image.png
>
>
> We observed that when a NameNode becomes UNAVAILABLE, the corresponding
> blockpool id in MembershipStoreImpl#activeNamespaces on dfsrouter
> unintentionally sets to empty, its initial value.
> !image.png|height=250!
> As a result of this, concat operations through dfsrouter fail with the
> following error as it cannot resolve the block id in the recognized active
> namespaces.
> {noformat}
> Caused by:
> org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.ipc.RemoteException):
> Cannot locate a nameservice for block pool BP-...
> {noformat}
> A possible fix is to ignore UNAVAILABLE NameNode registrations, and set
> proper namespace information obtained from available NameNode registrations
> when constructing the cache of active namespaces.
>
> [https://github.com/apache/hadoop/blob/rel/release-3.3.0/hadoop-hdfs-project/hadoop-hdfs-rbf/src/main/java/org/apache/hadoop/hdfs/server/federation/store/impl/MembershipStoreImpl.java#L207-L221]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
