[
https://issues.apache.org/jira/browse/HADOOP-4885?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12667512#action_12667512
]
Boris Shkolnik commented on HADOOP-4885:
----------------------------------------
other suggestions for simulating storage failures in the test:
1. Using Security Manager
unfortunately security manager is consulted during opening of a file for
writing, but not for a writing into already open stream.
2. deleting directories and calling doCheckpoint to force creation of a new
edits file in these directories (which will fail and cause the storage to be
moved to removedStorages).
This looks promising. I am going to investigate this approach.
> Try to restore failed replicas of Name Node storage (at checkpoint time)
> ------------------------------------------------------------------------
>
> Key: HADOOP-4885
> URL: https://issues.apache.org/jira/browse/HADOOP-4885
> Project: Hadoop Core
> Issue Type: Improvement
> Components: dfs
> Reporter: Boris Shkolnik
> Assignee: Boris Shkolnik
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
