hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Todd Lipcon <t...@cloudera.com>
Subject Re: Region Servers Crashing during Random Reads
Date Thu, 03 Feb 2011 21:43:35 GMT
Hi Charan,

Your GC settings are way off - 6m newsize will promote way too much to the
oldgen.

Try this:

-XX:+UseConcMarkSweepGC -XX:+UseParNewGC -Xmn256m
-XX:CMSInitiatingOccupancyFraction=70

-Todd

On Thu, Feb 3, 2011 at 12:28 PM, charan kumar <charan.kumar@gmail.com>wrote:

> HI Jonathan,
>
>  Thanks for you quick reply..
>
> Heap is set to 4G.
>
> Following are the JVM opts.
> export HBASE_OPTS="$HBASE_OPTS -XX:+HeapDumpOnOutOfMemoryError
> -XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode -XX:NewSize=6m
> -XX:MaxNewSize=6m"
>
> Are there any other options apart from increasing the RAM?
>
> I am adding some more info about the app.
>
>  > We are storing web page data in HBase.
>  > Row key is Hashed URL, for random distribution, since we dont plan to do
> scan's..
>  > We have LZOCompression Set on this column family.
>  > We were noticing 1500 Reads, when reading the page content.
>  > We have a column family, which stores just metadata of the page "title"
> etc... When reading this the performance is whopping 12000 TPS.
>
>  We though the issue could be because of N/w bandwidth used between HBase
> and Clients. So we disable LZO Compression on Column Family and started
> doing the compression of the raw page on the client and decompress it when
> readind (LZO).
>
>  > With this my write performance jumped up from 2000 to 5000 at peak.
>  > With this approach, the servers are crashing... Not sure , why only
> after
> turning of LZO... and doing the same from client.
>
>
>
> On Thu, Feb 3, 2011 at 12:13 PM, Jonathan Gray <jgray@fb.com> wrote:
>
> > How much heap are you running on your RegionServers?
> >
> > 6GB of total RAM is on the low end.  For high throughput applications, I
> > would recommend at least 6-8GB of heap (so 8+ GB of RAM).
> >
> > > -----Original Message-----
> > > From: charan kumar [mailto:charan.kumar@gmail.com]
> > > Sent: Thursday, February 03, 2011 11:47 AM
> > > To: user@hbase.apache.org
> > > Subject: Region Servers Crashing during Random Reads
> > >
> > > Hello,
> > >
> > >  I am using hbase 0.90.0 with hadoop-append. h/w ( Dell 1950, 2 CPU, 6
> GB
> > > RAM)
> > >
> > > I had 9 Region Servers crash (out of 30) in a span of 30 minutes during
> a
> > heavy
> > > reads. It looks like a GC, ZooKeeper Connection Timeout thingy to me.
> > > I did all recommended configuration from the Hbase wiki... Any other
> > > suggestions?
> > >
> > >
> > > 2011-02-03T09:43:07.890-0800: 70693.632: [GC 70693.632: [ParNew
> > > (promotion
> > > failed): 5555K->5540K(5568K), 0.0280950 secs]70693.660:
> > > [CMS2011-02-03T09:43:16.864-0800: 70702.606: [CMS-concurrent-mark:
> > > 12.549/69.323 secs] [Times: user=11.90 sys=1.26, real=69.31 secs]
> > >
> > > 2011-02-03T09:53:35.165-0800: 71320.785: [GC 71320.785: [ParNew
> > > (promotion
> > > failed): 5568K->5568K(5568K), 0.4384530 secs]71321.224:
> > > [CMS2011-02-03T09:53:45.111-0800: 71330.731: [CMS-concurrent-mark:
> > > 17.511/51.564 secs] [Times: user=38.72 sys=5.67, real=51.60 secs]
> > >
> > > 2011-02-03T09:43:07.890-0800: 70693.632: [GC 70693.632: [ParNew
> > > (promotion
> > > failed): 5555K->5540K(5568K), 0.0280950 secs]70693.660:
> > > [CMS2011-02-03T09:43:16.864-0800: 70702.606: [CMS-concurrent-mark:
> > > 12.549/69.323 secs] [Times: user=11.90 sys=1.26, real=69.31 secs]
> > >
> > >
> > > The following is the log entry in region Server
> > >
> > > 2011-02-03 10:37:43,946 INFO org.apache.zookeeper.ClientCnxn: Client
> > > session timed out, have not heard from server in 47172ms for sessionid
> > > 0x12db9f722421ce3, closing socket connection and attempting reconnect
> > > 2011-02-03 10:37:43,947 INFO org.apache.zookeeper.ClientCnxn: Client
> > > session timed out, have not heard from server in 48159ms for sessionid
> > > 0x22db9f722501d93, closing socket connection and attempting reconnect
> > > 2011-02-03 10:37:44,401 INFO org.apache.zookeeper.ClientCnxn: Opening
> > > socket connection to server XXXXXXXXXXXXXXXX
> > > 2011-02-03 10:37:44,402 INFO org.apache.zookeeper.ClientCnxn: Socket
> > > connection established to XXXXXXXXX, initiating session
> > > 2011-02-03 10:37:44,709 INFO org.apache.zookeeper.ClientCnxn: Opening
> > > socket connection to server XXXXXXXXXXXXXXX
> > > 2011-02-03 10:37:44,709 INFO org.apache.zookeeper.ClientCnxn: Socket
> > > connection established to XXXXXXXXXXXXXXXXXXXXX, initiating session
> > > 2011-02-03 10:37:44,767 DEBUG
> > > org.apache.hadoop.hbase.io.hfile.LruBlockCache: Block cache LRU
> eviction
> > > started; Attempting to free 81.93 MB of total=696.25 MB
> > > 2011-02-03 10:37:44,784 DEBUG
> > > org.apache.hadoop.hbase.io.hfile.LruBlockCache: Block cache LRU
> eviction
> > > completed; freed=81.94 MB, total=614.81 MB, single=379.98 MB,
> > > multi=309.77 MB, memory=0 KB
> > > 2011-02-03 10:37:45,205 INFO org.apache.zookeeper.ClientCnxn: Unable to
> > > reconnect to ZooKeeper service, session 0x22db9f722501d93 has expired,
> > > closing socket connection
> > > 2011-02-03 10:37:45,206 INFO
> > > org.apache.hadoop.hbase.client.HConnectionManager$HConnectionImplem
> > > entation:
> > > This client just lost it's session with ZooKeeper, trying to reconnect.
> > > 2011-02-03 10:37:45,453 INFO
> > > org.apache.hadoop.hbase.client.HConnectionManager$HConnectionImplem
> > > entation:
> > > Trying to reconnect to zookeeper
> > > 2011-02-03 10:37:45,206 INFO org.apache.zookeeper.ClientCnxn: Unable to
> > > reconnect to ZooKeeper service, session 0x12db9f722421ce3 has expired,
> > > closing socket connection
> > > gionserver:60020-0x22db9f722501d93 regionserver:60020-
> > > 0x22db9f722501d93
> > > received expired from ZooKeeper, aborting
> > > org.apache.zookeeper.KeeperException$SessionExpiredException:
> > > KeeperErrorCode = Session expired
> > >         at
> > > org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher.connectionEvent(
> > > ZooKeeperWatcher.java:328)
> > >         at
> > > org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher.process(ZooKeep
> > > erWatcher.java:246)
> > >         at
> > > org.apache.zookeeper.ClientCnxn$EventThread.processEvent(ClientCnxn.ja
> > > va:530)
> > >         at
> > > org.apache.zookeeper.ClientCnxn$EventThread.run(ClientCnxn.java:506)
> > > handled exception: org.apache.hadoop.hbase.YouAreDeadException: Server
> > > REPORT rejected; currently processing XXXXXXXXXXXX,60020,1296684296172
> > > as dead server
> > > org.apache.hadoop.hbase.YouAreDeadException:
> > > org.apache.hadoop.hbase.YouAreDeadException: Server REPORT rejected;
> > > currently processing XXXXXXXXXXXX,60020,1296684296172 as dead server
> > >         at
> sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native
> > > Method)
> > >         at
> > > sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructor
> > > AccessorImpl.java:39)
> > >         at
> > > sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingCon
> > > structorAccessorImpl.java:27)
> > >         at
> > java.lang.reflect.Constructor.newInstance(Constructor.java:513)
> > >         at
> > > org.apache.hadoop.ipc.RemoteException.instantiateException(RemoteExce
> > > ption.java:96)
> > >         at
> > > org.apache.hadoop.ipc.RemoteException.unwrapRemoteException(Remote
> > > Exception.java:80)
> > >         at
> > > org.apache.hadoop.hbase.regionserver.HRegionServer.tryRegionServerRep
> > > ort(HRegionServer.java:729)
> > >         at
> > > org.apache.hadoop.hbase.regionserver.HRegionServer.run(HRegionServer.j
> > > ava:586)
> > >         at java.lang.Thread.run(Thread.java:619)
> > >
> > >
> > > 2011-02-03T09:53:35.165-0800: 71320.785: [GC 71320.785: [ParNew
> > > (promotion
> > > failed): 5568K->5568K(5568K), 0.4384530 secs]71321.224:
> > > [CMS2011-02-03T09:53:45.111-0800: 71330.731: [CMS-concurrent-mark:
> > > 17.511/51.564 secs] [Times: user=38.72 sys=5.67, real=51.60 secs]
> > >
> > >
> > >
> > > Thanks,
> > > Charan
> >
>



-- 
Todd Lipcon
Software Engineer, Cloudera

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message