hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Liu, Raymond" <raymond....@intel.com>
Subject RE: One weird problem of my MR job upon hbase table.
Date Mon, 14 Jan 2013 08:26:28 GMT

For feedback.

With a lot of profiling works, I guess I found the most promise cause of my problem.

It's not because one disk is slow or something ( though I do have slow disk on different region
servers, but the lagging behind pattern seems not related to the disk slowness pattern)

It seems to me this issue is caused by HDFS blocks not distributed evenly across disks on
the same region server.

For a specific table, I do have even region across region servers. But the HFILE data blocks
that belong to the local region on the slow region server don't distributed evenly. Say disk1
have 45 blocks, while disk4 have 30 blocks etc.

While, on the large, Add up all the blocks from all tables including both local region's block
and remote region's replica data block on this region server, they are evenly distributed.

Thus I guess the HDFS did try to even out the data block on the disks, But since it do not
know which block is belong to which region, and there are incoming replica data blocks, so
even with round robin strategy, it could not even out the "local region"'s data block across
disks. Seems this could hardly been avoid?

Thus, there are hotspot disk, and with defined scan sequence which lead to lag behind region,
which lead to lag behind map task and on this region server.

This is the best guess I have gain up to now. But not knowing why this issue come out suddenly
on my cluster, or why I don't observe it before...

> Hi there,
> The HBase RefGuide has a comprehensive case study on such a case.  This
> might not be the exact problem, but the diagnostic approach should help.
> http://hbase.apache.org/book.html#casestudies.slownode
> On 1/4/13 10:37 PM, "Liu, Raymond" <raymond.liu@intel.com> wrote:
> >Hi
> >
> >I encounter a weird lag behind map task issue here :
> >
> >I have a small hadoop/hbase cluster with 1 master node and 4
> >regionserver node all have 16 CPU with map and reduce slot set to 24.
> >
> >A few table is created with regions distributed on each region node
> >evenly ( say 16 region for each region server). Also each region has
> >almost the same number of kvs with very similar size. All table had
> >major_compact done to ensure data locality
> >
> >I have a MR job which simply do local region scan in every map task (
> >so
> >16 map task for each regionserver node).
> >
> >By theory, every map task should finish within similar time.
> >
> >But the real case is that some regions on the same region server always
> >lags behind a lot, say cost 150 ~250% of the other map tasks average
> >times.
> >
> >If this is happen to a single region server for every table, I might
> >doubt it is a disk issue or other reason that bring down the
> >performance of this region server.
> >
> >But the weird thing is that, though with each single table, almost all
> >the map task on the the same single regionserver is lag behind. But for
> >different table, this lag behind regionserver is different! And the
> >region and region size is distributed evenly which I double checked for
> >a lot of times. ( I even try to set replica to 4 to ensure every node
> >have a copy of local data)
> >
> >Say table 1, all map task on regionserver node 2 is slow. While for
> >table 2, maybe all map task on regionserver node 3 is slow, and with
> >table 1, it will always be regionserver node 2 which is slow regardless
> >of cluster restart, and the slowest map task will always be the very
> >same one. And it won't go away even I do major compact again.....
> >
> >So, anyone could give me some clue on what reason might possible lead
> >to this weird behavior? Any wild guess is welcome!
> >
> >(BTW. I don't encounter this issue a few days ago with the same table.
> >While I do restart cluster and do a few changes upon config file during
> >that period, But restore the config file don't help)
> >
> >
> >Best Regards,
> >Raymond Liu
> >
> >

View raw message