[
https://issues.apache.org/jira/browse/HADOOP-4584?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12697149#action_12697149
]
Raghu Angadi commented on HADOOP-4584:
--------------------------------------
Mostly looks good. A few comments :
# default scan period is one hour (same as before).. I think it should be much
less often (may be 6 to 24 hours).
# Since there is no throttling of directory scan, it is better to randomize the
start time. The datanodes are usually started at the same time, the whole
cluster could slow down at the same time.
# regex patterns for block files : '.' in ".meta" needs to be escaped. Need a
'$' at the end for more accuracy?
# diskGS in checkAndUpdate() is calculated outside lock. Any correlation is
valid only inside the lock. Could be a problem now or later.
# At patchfile:834 : It updates generation stamp with 'diskGS' without moving
the meta file from prev directory to memBlock's directory. Could that result in
block and meta files in different directories?
> Slow generation of blockReport at DataNode causes delay of sending heartbeat
> to NameNode
> ----------------------------------------------------------------------------------------
>
> Key: HADOOP-4584
> URL: https://issues.apache.org/jira/browse/HADOOP-4584
> Project: Hadoop Core
> Issue Type: Bug
> Components: dfs
> Reporter: Hairong Kuang
> Assignee: Suresh Srinivas
> Fix For: 0.20.0
>
> Attachments: 4584.brthread.2.patch, 4584.brthread.3.patch,
> 4584.brthread.3.patch, 4584.brthread.3.patch, 4584.brthread.3.patch,
> 4584.brthread.3.patch, 4584.brthread.4.patch, 4584.brthread.4.patch,
> 4584.brthread.4.patch, 4584.hbthread.patch, 4584.patch, 4584.patch,
> 4584.patch, 4584.patch, 4584.patch, 4584.patch, Design.pdf, Design.pdf
>
>
> sometimes due to disk or some other problems, datanode takes minutes or tens
> of minutes to generate a block report. It causes the datanode not able to
> send heartbeat to NameNode every 3 seconds. In the worst case, it makes
> NameNode to detect a lost heartbeat and wrongly decide that the datanode is
> dead.
> It would be nice to have two threads instead. One thread is for scanning data
> directories and generating block report, and executes the requests sent by
> NameNode; Another thread is for sending heartbeats, block reports, and
> picking up the requests from NameNode. By having these two threads, the
> sending of heartbeats will not get delayed by any slow block report or slow
> execution of NameNode requests.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
