Re: Preventing SessionExpired events
Hello Mahadev, Thanks for the tip on GC parameters. I will try this. I have increased the HBase default 10 second session timeout to 60 seconds, but will try the GC config suggested here and begin reducing the timeout to smaller values to determine its effect on the issues I have been encountering. Thanks again, - Andy From: Mahadev Konar Subject: Re: Preventing SessionExpired events Date: Wednesday, April 8, 2009, 1:40 PM Nitay, Thanks for sending us the info. We have experienced such gc problem in our HDFS (hadoop file system) setups. The gc had been quite a problem for us with the Namenode (hadoop hdfs) process. We have seen the namenode just stalling for minutes doing garbage collection. We currently run the namenode with the following gc options -XX:ParallelGCThreads=8 -XX:+UseConcMarkSweepGC And have avoided getting into trouble with the gc.
Re: Preventing SessionExpired events
Also, I should mention that some of the errors Andrew was seeing are related to ZOOKEEPER-344: I see this kind of stuff: 2009-04-07 17:58:13,344 - WARN [NIOServerCxn.Factory:2181: nioserverc...@417] - Exception causing close of session 0x2208296c38e due to java.io.IOException: Read error and bye bye HRS ephemeral znodes, which triggers (currently) HBASE-1314. This I think is ZOOKEEPER-344 https://issues.apache.org/jira/browse/ZOOKEEPER-344 - Andy On Wed, Apr 8, 2009 at 12:39 PM, Nitay nit...@gmail.com wrote: Hey guys, We've recently replaced a few pieces of HBase's cluster management and coordination with ZooKeeper. One of guys, Andrew Purtell, has a cluster that he throws a lot of load at. Andrew's cluster was getting a lot of SessionExpired events which were causing some havoc. After some discussion on the hbase list and additional testing by Andrew (tweaking things like the session timeout, quorum size, and GC used), we suspect the problem is that the Java GC is starving the ZooKeeper hearbeat thread from executing. There is a JIRA open on the matter where Joey suggests a solution that has worked for him: https://issues.apache.org/jira/browse/HBASE-1316 We wanted to loop you guys in to see if you have any thoughts/suggestions on the matter. Thanks, -n
Re: Preventing SessionExpired events
What are you running for a session timeout on your clients? Can you run with something like jvisualvm or jconsole, and watch the gc activity when the session timeouts occur? Might give you some insight. Have you tried one of the alternative GC's available in the VM? http://developer.amd.com/documentation/articles/pages/4EasyWaystodoJavaGarbageCollectionTuning.aspx ie Flags for Latency Applications We are also working on the following jira: https://issues.apache.org/jira/browse/ZOOKEEPER-321 which will eliminate session expirations for clients w/o ephemerals. (is this the case for you?) Try turning on debug in your client, the client will spit out: LOG.debug(Got ping response for sessionid:0x If you turn on trace logging in the server you should see session updates there as well (c-server, which control session expiration). re HBASE-1316 - how does the jni c wrapper fix this? Isn't the code still running w/in the same (vm) process? Unfortunately I can't think of anything else if it is the GC. Basically you'd have to increase the timeout or try another gc with lower latency. Perhaps Mahadev/Ben/Flavio might have insight... Patrick Nitay wrote: Hey guys, We've recently replaced a few pieces of HBase's cluster management and coordination with ZooKeeper. One of guys, Andrew Purtell, has a cluster that he throws a lot of load at. Andrew's cluster was getting a lot of SessionExpired events which were causing some havoc. After some discussion on the hbase list and additional testing by Andrew (tweaking things like the session timeout, quorum size, and GC used), we suspect the problem is that the Java GC is starving the ZooKeeper hearbeat thread from executing. There is a JIRA open on the matter where Joey suggests a solution that has worked for him: https://issues.apache.org/jira/browse/HBASE-1316 We wanted to loop you guys in to see if you have any thoughts/suggestions on the matter. Thanks, -n
Re: Preventing SessionExpired events
This is good to know. It will allow us to try an replicate the situation, which we haven't been able to do. I'm hoping we can come up with something that we can proactively do to address this... Patrick Nitay wrote: Also, I should mention that some of the errors Andrew was seeing are related to ZOOKEEPER-344: I see this kind of stuff: 2009-04-07 17:58:13,344 - WARN [NIOServerCxn.Factory:2181: nioserverc...@417] - Exception causing close of session 0x2208296c38e due to java.io.IOException: Read error and bye bye HRS ephemeral znodes, which triggers (currently) HBASE-1314. This I think is ZOOKEEPER-344 https://issues.apache.org/jira/browse/ZOOKEEPER-344 - Andy On Wed, Apr 8, 2009 at 12:39 PM, Nitay nit...@gmail.com wrote: Hey guys, We've recently replaced a few pieces of HBase's cluster management and coordination with ZooKeeper. One of guys, Andrew Purtell, has a cluster that he throws a lot of load at. Andrew's cluster was getting a lot of SessionExpired events which were causing some havoc. After some discussion on the hbase list and additional testing by Andrew (tweaking things like the session timeout, quorum size, and GC used), we suspect the problem is that the Java GC is starving the ZooKeeper hearbeat thread from executing. There is a JIRA open on the matter where Joey suggests a solution that has worked for him: https://issues.apache.org/jira/browse/HBASE-1316 We wanted to loop you guys in to see if you have any thoughts/suggestions on the matter. Thanks, -n
Re: Preventing SessionExpired events
The default session timeout in HBase is currently 10 seconds. Bumping it up to 30 and 60 reduced SessionExpired exceptions, according to Andrew. I believe Andrew did run it under jconsole. He was also tuning GC parameters. He mentioned running using incremental garbage collector (-XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode). He can provide more details on all of this. My understanding with HBASE-1316 is that it solves the problem because the ZooKeeper IO/hearbeat thread becomes an OS level thread which is not managed by Java. Hence, the GC does not starve it. Joey can comment here as he developed the solution. There are three main components that use ZooKeeper in HBase are: client, regionserver, and master. The client does not have ephemeral nodes so having something like ZOOKEEPER-321 for it would be nice. It is currently read only. For now recovering it by reinitializing the ZooKeeper handle is not a big deal. The bigger issue is with the master and regionserver, which do use ephemeral nodes. Recovering them is a bit tougher, and we'd like to prevent getting SessionExpired as much as possible. On Wed, Apr 8, 2009 at 1:17 PM, Patrick Hunt ph...@apache.org wrote: What are you running for a session timeout on your clients? Can you run with something like jvisualvm or jconsole, and watch the gc activity when the session timeouts occur? Might give you some insight. Have you tried one of the alternative GC's available in the VM? http://developer.amd.com/documentation/articles/pages/4EasyWaystodoJavaGarbageCollectionTuning.aspx ie Flags for Latency Applications We are also working on the following jira: https://issues.apache.org/jira/browse/ZOOKEEPER-321 which will eliminate session expirations for clients w/o ephemerals. (is this the case for you?) Try turning on debug in your client, the client will spit out: LOG.debug(Got ping response for sessionid:0x If you turn on trace logging in the server you should see session updates there as well (c-server, which control session expiration). re HBASE-1316 - how does the jni c wrapper fix this? Isn't the code still running w/in the same (vm) process? Unfortunately I can't think of anything else if it is the GC. Basically you'd have to increase the timeout or try another gc with lower latency. Perhaps Mahadev/Ben/Flavio might have insight... Patrick Nitay wrote: Hey guys, We've recently replaced a few pieces of HBase's cluster management and coordination with ZooKeeper. One of guys, Andrew Purtell, has a cluster that he throws a lot of load at. Andrew's cluster was getting a lot of SessionExpired events which were causing some havoc. After some discussion on the hbase list and additional testing by Andrew (tweaking things like the session timeout, quorum size, and GC used), we suspect the problem is that the Java GC is starving the ZooKeeper hearbeat thread from executing. There is a JIRA open on the matter where Joey suggests a solution that has worked for him: https://issues.apache.org/jira/browse/HBASE-1316 We wanted to loop you guys in to see if you have any thoughts/suggestions on the matter. Thanks, -n
Re: Preventing SessionExpired events
Nitay, Thanks for sending us the info. We have experienced such gc problem in our HDFS (hadoop file system) setups. The gc had been quite a problem for us with the Namenode (hadoop hdfs) process. We have seen the namenode just stalling for minutes doing garbage collection. We currently run the namenode with the following gc options -XX:ParallelGCThreads=8 -XX:+UseConcMarkSweepGC And have avoided getting into trouble with the gc. What options are you guys using with the java process that embeds the java zookeeper client? Maybe the above gc options could help. Also, we would be interested in having such jni wrappers into the c libarary in case more people want it. Is the wrapper posted somewhere for us to take a look at? Thanks mahadev On 4/8/09 1:17 PM, Patrick Hunt ph...@apache.org wrote: What are you running for a session timeout on your clients? Can you run with something like jvisualvm or jconsole, and watch the gc activity when the session timeouts occur? Might give you some insight. Have you tried one of the alternative GC's available in the VM? http://developer.amd.com/documentation/articles/pages/4EasyWaystodoJavaGarbage CollectionTuning.aspx ie Flags for Latency Applications We are also working on the following jira: https://issues.apache.org/jira/browse/ZOOKEEPER-321 which will eliminate session expirations for clients w/o ephemerals. (is this the case for you?) Try turning on debug in your client, the client will spit out: LOG.debug(Got ping response for sessionid:0x If you turn on trace logging in the server you should see session updates there as well (c-server, which control session expiration). re HBASE-1316 - how does the jni c wrapper fix this? Isn't the code still running w/in the same (vm) process? Unfortunately I can't think of anything else if it is the GC. Basically you'd have to increase the timeout or try another gc with lower latency. Perhaps Mahadev/Ben/Flavio might have insight... Patrick Nitay wrote: Hey guys, We've recently replaced a few pieces of HBase's cluster management and coordination with ZooKeeper. One of guys, Andrew Purtell, has a cluster that he throws a lot of load at. Andrew's cluster was getting a lot of SessionExpired events which were causing some havoc. After some discussion on the hbase list and additional testing by Andrew (tweaking things like the session timeout, quorum size, and GC used), we suspect the problem is that the Java GC is starving the ZooKeeper hearbeat thread from executing. There is a JIRA open on the matter where Joey suggests a solution that has worked for him: https://issues.apache.org/jira/browse/HBASE-1316 We wanted to loop you guys in to see if you have any thoughts/suggestions on the matter. Thanks, -n
Re: Preventing SessionExpired events
Nitay is correct about the native threads. Using the pure Java API, the garbage collector will occasionally pause other Java threads to do a full mark and sweep. Even switching to the concurrent collector only delays the problem. The issues is mixing a high throughput application (HBase) with a low latency library (Zookeeper). Systems like HBase live on relatively large numbers of short lived objects. You only key keys and values long enough for the Memcache to get full then you write all the data to HDFS and throw away the objects. You can patch around the issue with object pools, but ultimately you need to insulate zk from the GC pauses. In our experience, the best way to do that was a jni wrapper around the zk C api. Since the C api uses it's own posix threads, it's protected from the GC. In the system we wrote, we ended up using the Java api with a large session timeout for most everything, and used the jni code just for creating ephemeral nodes. -Joey On Wed, Apr 8, 2009 at 9:35 PM, Nitay nit...@gmail.com wrote: The default session timeout in HBase is currently 10 seconds. Bumping it up to 30 and 60 reduced SessionExpired exceptions, according to Andrew. I believe Andrew did run it under jconsole. He was also tuning GC parameters. He mentioned running using incremental garbage collector (-XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode). He can provide more details on all of this. My understanding with HBASE-1316 is that it solves the problem because the ZooKeeper IO/hearbeat thread becomes an OS level thread which is not managed by Java. Hence, the GC does not starve it. Joey can comment here as he developed the solution. There are three main components that use ZooKeeper in HBase are: client, regionserver, and master. The client does not have ephemeral nodes so having something like ZOOKEEPER-321 for it would be nice. It is currently read only. For now recovering it by reinitializing the ZooKeeper handle is not a big deal. The bigger issue is with the master and regionserver, which do use ephemeral nodes. Recovering them is a bit tougher, and we'd like to prevent getting SessionExpired as much as possible. On Wed, Apr 8, 2009 at 1:17 PM, Patrick Hunt ph...@apache.org wrote: What are you running for a session timeout on your clients? Can you run with something like jvisualvm or jconsole, and watch the gc activity when the session timeouts occur? Might give you some insight. Have you tried one of the alternative GC's available in the VM? http://developer.amd.com/documentation/articles/pages/4EasyWaystodoJavaGarbageCollectionTuning.aspx ie Flags for Latency Applications We are also working on the following jira: https://issues.apache.org/jira/browse/ZOOKEEPER-321 which will eliminate session expirations for clients w/o ephemerals. (is this the case for you?) Try turning on debug in your client, the client will spit out: LOG.debug(Got ping response for sessionid:0x If you turn on trace logging in the server you should see session updates there as well (c-server, which control session expiration). re HBASE-1316 - how does the jni c wrapper fix this? Isn't the code still running w/in the same (vm) process? Unfortunately I can't think of anything else if it is the GC. Basically you'd have to increase the timeout or try another gc with lower latency. Perhaps Mahadev/Ben/Flavio might have insight... Patrick Nitay wrote: Hey guys, We've recently replaced a few pieces of HBase's cluster management and coordination with ZooKeeper. One of guys, Andrew Purtell, has a cluster that he throws a lot of load at. Andrew's cluster was getting a lot of SessionExpired events which were causing some havoc. After some discussion on the hbase list and additional testing by Andrew (tweaking things like the session timeout, quorum size, and GC used), we suspect the problem is that the Java GC is starving the ZooKeeper hearbeat thread from executing. There is a JIRA open on the matter where Joey suggests a solution that has worked for him: https://issues.apache.org/jira/browse/HBASE-1316 We wanted to loop you guys in to see if you have any thoughts/suggestions on the matter. Thanks, -n
RE: Preventing SessionExpired events
I'm curious about your session scenario. If your servers can hang for X seconds without a problem and Y is your session timeout, why would you set Y X? In your case if you set your session timeout to 5 secs for example, but your server can hang for 20 seconds doing GC, your clients cannot expect a response of less than 20 seconds, so why don't you set your session timeout to 20 seconds? ben From: Joey Echeverria [joe...@gmail.com] Sent: Wednesday, April 08, 2009 1:47 PM To: hbase-...@hadoop.apache.org Cc: zookeeper-dev@hadoop.apache.org Subject: Re: Preventing SessionExpired events Nitay is correct about the native threads. Using the pure Java API, the garbage collector will occasionally pause other Java threads to do a full mark and sweep. Even switching to the concurrent collector only delays the problem. The issues is mixing a high throughput application (HBase) with a low latency library (Zookeeper). Systems like HBase live on relatively large numbers of short lived objects. You only key keys and values long enough for the Memcache to get full then you write all the data to HDFS and throw away the objects. You can patch around the issue with object pools, but ultimately you need to insulate zk from the GC pauses. In our experience, the best way to do that was a jni wrapper around the zk C api. Since the C api uses it's own posix threads, it's protected from the GC. In the system we wrote, we ended up using the Java api with a large session timeout for most everything, and used the jni code just for creating ephemeral nodes. -Joey On Wed, Apr 8, 2009 at 9:35 PM, Nitay nit...@gmail.com wrote: The default session timeout in HBase is currently 10 seconds. Bumping it up to 30 and 60 reduced SessionExpired exceptions, according to Andrew. I believe Andrew did run it under jconsole. He was also tuning GC parameters. He mentioned running using incremental garbage collector (-XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode). He can provide more details on all of this. My understanding with HBASE-1316 is that it solves the problem because the ZooKeeper IO/hearbeat thread becomes an OS level thread which is not managed by Java. Hence, the GC does not starve it. Joey can comment here as he developed the solution. There are three main components that use ZooKeeper in HBase are: client, regionserver, and master. The client does not have ephemeral nodes so having something like ZOOKEEPER-321 for it would be nice. It is currently read only. For now recovering it by reinitializing the ZooKeeper handle is not a big deal. The bigger issue is with the master and regionserver, which do use ephemeral nodes. Recovering them is a bit tougher, and we'd like to prevent getting SessionExpired as much as possible. On Wed, Apr 8, 2009 at 1:17 PM, Patrick Hunt ph...@apache.org wrote: What are you running for a session timeout on your clients? Can you run with something like jvisualvm or jconsole, and watch the gc activity when the session timeouts occur? Might give you some insight. Have you tried one of the alternative GC's available in the VM? http://developer.amd.com/documentation/articles/pages/4EasyWaystodoJavaGarbageCollectionTuning.aspx ie Flags for Latency Applications We are also working on the following jira: https://issues.apache.org/jira/browse/ZOOKEEPER-321 which will eliminate session expirations for clients w/o ephemerals. (is this the case for you?) Try turning on debug in your client, the client will spit out: LOG.debug(Got ping response for sessionid:0x If you turn on trace logging in the server you should see session updates there as well (c-server, which control session expiration). re HBASE-1316 - how does the jni c wrapper fix this? Isn't the code still running w/in the same (vm) process? Unfortunately I can't think of anything else if it is the GC. Basically you'd have to increase the timeout or try another gc with lower latency. Perhaps Mahadev/Ben/Flavio might have insight... Patrick Nitay wrote: Hey guys, We've recently replaced a few pieces of HBase's cluster management and coordination with ZooKeeper. One of guys, Andrew Purtell, has a cluster that he throws a lot of load at. Andrew's cluster was getting a lot of SessionExpired events which were causing some havoc. After some discussion on the hbase list and additional testing by Andrew (tweaking things like the session timeout, quorum size, and GC used), we suspect the problem is that the Java GC is starving the ZooKeeper hearbeat thread from executing. There is a JIRA open on the matter where Joey suggests a solution that has worked for him: https://issues.apache.org/jira/browse/HBASE-1316 We wanted to loop you guys in to see if you have any thoughts/suggestions on the matter. Thanks, -n
