[
https://issues.apache.org/jira/browse/HDFS-6574?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Liang Xie updated HDFS-6574:
----------------------------
Description:
My colleague [~cuijianwei] found in a HBase testing scenario, once a bad disk
occured, the local read will be skipped and lots of remote reads be requested
for a lengthy time, say, tens of minutes, then we had to trigger a compaction
to help recovering the locality and read latency.
It turned out relating with the addToDeadNodes(), imaging a disk in local node
has something wrong, current impl will add the entity local node to the dead
node list, then all other good disks in local node could not get read request
any more.
So better choices here to me, seems:
1) tell the detail IOException really is a connection related exception, then
call addToDeadNodes(). or
2) tell the IOException is related with bad block/disk, w/o call
addToDeadNodes(); else call addToDeadNodes().
another thing need to consider is if we have got a disk exception from one
node, should we refresh the locatedBlocks info from nn to clear all rotten
caching for that bad disk of the node ? it'll be heavy somehow if it's a huge
size file...
We have a plan to make a patch soon for our internal hadoop branch, due to
it'll degrade HBase read performance severely once a sick disk ocurred, also
we'd like to contribute to community if you think this is not too crazy...
[~stack]
was:
My colleague [~cuijianwei] found in a HBase testing scenario, once a bad disk
occured, the local read will be skipped and lots of remote read be requested
for a lengthy time, say, tens of minutes, then we had to trigger a compaction
to help recovering the locality and read latency.
It turned out relating with the addToDeadNodes(), imaging one disk in local
node has something wrong, current impl will add the local node to the dead node
list, then all other good disks in local node could not service any read
request.
So better chooses here to me, seems:
1) tell the detail IOException really is a connection related exception, then
call addToDeadNodes().
2) tell the IOException is related with bad block/disk, w/o call
addToDeadNodes(); else call addToDeadNodes().
another thing need to consider is if we have got a disk exception from one
node, should we refresh the locatedBlocks info from nn to clear all rotten
caching for that bad disk of the node ? it'll be heavy somehow if it's a huge
size file...
I have a plan to make a patch for our internal hadoop branch, due to it'll
degrade HBase read performance severely, also i'd like to contribute to
community if you think this proposal is not too crazy... [~stack]
> make sure addToDeadNodes() only be called when there's a connection issue
> occur
> -------------------------------------------------------------------------------
>
> Key: HDFS-6574
> URL: https://issues.apache.org/jira/browse/HDFS-6574
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: hdfs-client
> Affects Versions: 3.0.0, 2.5.0
> Reporter: Liang Xie
> Assignee: Liang Xie
>
> My colleague [~cuijianwei] found in a HBase testing scenario, once a bad disk
> occured, the local read will be skipped and lots of remote reads be requested
> for a lengthy time, say, tens of minutes, then we had to trigger a compaction
> to help recovering the locality and read latency.
> It turned out relating with the addToDeadNodes(), imaging a disk in local
> node has something wrong, current impl will add the entity local node to the
> dead node list, then all other good disks in local node could not get read
> request any more.
> So better choices here to me, seems:
> 1) tell the detail IOException really is a connection related exception, then
> call addToDeadNodes(). or
> 2) tell the IOException is related with bad block/disk, w/o call
> addToDeadNodes(); else call addToDeadNodes().
> another thing need to consider is if we have got a disk exception from one
> node, should we refresh the locatedBlocks info from nn to clear all rotten
> caching for that bad disk of the node ? it'll be heavy somehow if it's a
> huge size file...
> We have a plan to make a patch soon for our internal hadoop branch, due to
> it'll degrade HBase read performance severely once a sick disk ocurred, also
> we'd like to contribute to community if you think this is not too crazy...
> [~stack]
--
This message was sent by Atlassian JIRA
(v6.2#6252)
