Terracotta Discussion Forums

rajoshi wrote:

Hi ,

Can you provide complete client/server logs from the time of issue. Also detail of your set up (How many clients /server)

the server run 2-3 hours..

here is 8 wasphere 7.0 + tc server

2012-07-06 09:59:17,211 INFO - Terracotta 3.4.0, as of 20101104-141156 (Revision 16614 by cruise@su10mo4 from 3.4)
2012-07-06 09:59:17,898 INFO - Successfully loaded base configuration from file at '/home/ehr/tbCacheServer/bin/tc-config.xml'.
2012-07-06 09:59:18,157 INFO - Log file: '/home/ehr/terracotta/server-logs/terracotta-server.log'.
2012-07-06 09:59:20,846 INFO - Available Max Runtime Memory: 2254MB
2012-07-06 09:59:23,404 INFO - JMX Server started. Available at URL[service:jmx:jmxmp://0.0.0.0:9520]
2012-07-06 09:59:30,046 INFO - Becoming State[ ACTIVE-COORDINATOR ]
2012-07-06 09:59:30,119 INFO - Terracotta Server instance has started up as ACTIVE node on 0:0:0:0:0:0:0:0:9510 successfully, and is now ready for work.
2012-07-07 18:22:50,280 INFO - "Client Cannot Reconnect ConnectionID(1.94d8de597a35458b9deb8d31f0c71376) not found. Connection attempts from the Terracotta node at 10.233.72.22:43487 are being rejected by the Terracotta server array. Restart the Client to allow it to rejoin the cluster. Many Client reconnection failures can be avoided by configuring the Terracotta server array for "permanent-store" and tuning reconnection parameters. For more information, see http://www.terracotta.org/kit/reflector?kitID=default&pageID=HA"

thanks .later I will send the full log...

by the way ,the JDK version is 1.6.0_25..is it OK?

HI,hello:

I use the terracotta + ehcache..
The web Server is webSphere7.xxx
Now the was is locked...

thanks..please see the attach file

1、WAS7 and Terracotta miss the conection。
this is thd log.

2011-06-21 14:51:36,520 INFO - "Client Cannot Reconnect ConnectionID(8.6df43cb9ee84426aa177be3d472aea05) not found. Connection attempts from the Terracotta node at 10.105.19.166:20956 are being rejected by the Terracotta server array. Restart the Client to allow it to rejoin the cluster. Many Client reconnection failures can be avoided by configuring the Terracotta server array for "permanent-store" and tuning reconnection parameters. For more information, see http://www.terracotta.org/kit/reflector?kitID=default&pageID=HA"
2、the system project,was is locked?the log is
at sun.misc.Unsafe.park(Native Method)
at java.util.concurrent.locks.LockSupport.park(LockSupport.java:184)
at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:822)
at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:968)
at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1292)
at java.util.concurrent.Semaphore.acquire(Semaphore.java:297)
at com.tc.object.locks.LockStateNode$PendingLockHold.park(LockStateNode.java:183)
at com.tc.object.locks.ClientLockImpl.acquireQueued(ClientLockImpl.java:760)
at com.tc.object.locks.ClientLockImpl.acquireQueued(ClientLockImpl.java:748)
at com.tc.object.locks.ClientLockImpl.lock(ClientLockImpl.java:56)
at com.tc.object.locks.ClientLockManagerImpl.lock(ClientLockManagerImpl.java:99)
at com.tc.object.bytecode.ManagerImpl.lock(ManagerImpl.java:795)
at com.tc.object.bytecode.ManagerUtilInternal.beginLock(ManagerUtilInternal.java:56)
at org.terracotta.locking.strategy.LongLockStrategy.beginLock(LongLockStrategy.java:16)
at org.terracotta.locking.strategy.LongLockStrategy.beginLock(LongLockStrategy.java:7)
at com.terracotta.toolkit.collections.ConcurrentDistributedServerMapDso.beginLock(ConcurrentDistributedServerMapDso.java:168)
at com.terracotta.toolkit.collections.ConcurrentDistributedServerMapDso.get(ConcurrentDistributedServerMapDso.java:551)
at com.terracotta.toolkit.collections.ConcurrentDistributedMapDsoArray.get(ConcurrentDistributedMapDsoArray.java:217)
at org.terracotta.collections.ConcurrentDistributedMap.get(ConcurrentDistributedMap.java:212)
at org.terracotta.cache.TerracottaDistributedCache.getNonExpiredEntry(TerracottaDistributedCache.java:223)
at org.terracotta.cache.TerracottaDistributedCache.getNonExpiredEntryCoherent(TerracottaDistributedCache.java:152)
at org.terracotta.cache.TerracottaDistributedCache.getTimestampedValueQuiet(TerracottaDistributedCache.java:198)
at org.terracotta.modules.ehcache.store.ClusteredStoreBackendImpl.getTimestampedValueQuiet(ClusteredStoreBackendImpl.java:70)
at org.terracotta.modules.ehcache.store.backend.StrictBackend.get(StrictBackend.java:48)
at org.terracotta.modules.ehcache.store.ClusteredStore.doGet(ClusteredStore.java:370)
at org.terracotta.modules.ehcache.store.ClusteredStore.getQuiet(ClusteredStore.java:361)
at net.sf.ehcache.Cache.searchInStoreWithoutStats(Cache.java:1933)
at net.sf.ehcache.Cache.get(Cache.java:1555)
at net.sf.ehcache.Cache.get(Cache.java:1522)

thanks you..!

您好：
1台WAS作为客户段。
1台作为tc ehcache server

发现，有时会有缓存丢失
1 已经把ehcache中的缓存配成永久有效了。
2 如果就一台was，可以在put方法上加锁，解决并发，缓存不丢失？如果是集群发布was,如何作到不丢失节点呢？

多谢。

我自己试验了，很多参数，当用, terracotta监控，查看时。。

on-disk size 总是大于0的。

tc配置里面，<dso> </dso> 里面，用默认人temp-mode

请教。。如何，不往磁盘中写。我想写在内存中。因为大并发时，磁盘io 是个问题.

请老马给个电话，我想联系咨询一下。

在Linux系统下发布一个TC,两个机器是WAS7。大并发。查询。

另：想问请教一个别的问题。

关于，web容器与TC-server,时间同步问题，如果，不同步的话，有产生什么样的问题？
我现在，还没有同步。

经过测试

TC-server 一台 + WAS7 一台每秒并发数（150并发）

和

TC-server 一台 + WAS7 二台每秒并发数(300并发) 差不多。

这是不是，有问题，？感觉像是串行。

请问马老师和大家，有没有什么配置。

多谢老马的回复。

通过配置tc-config.xml。我已经调整为60秒钟一次DGC,可以解决我的问题，也看到了，在不断的更新过程时，当DGC运行的时候，garbage count确实有数字，收回来了垃圾。

默认的配置，我也修改了，改为，5000M,我在64位系统下。。但是这样，会不会，太频繁，有性能的问题呢？

另外，还有一个问题想请教。

免费版，可不可以使用。
1 server array,就是发布多个server,其他的server都是，stand-by.?
2 可不可以，把内存中的数据，存入磁盘，当server重启时，自动从磁盘中，载入缓存呢？

多谢。

如题，我在大并发或，在循环中，对一个缓存更新，这个对象大的小，线性增长。

1 这个时候，tc监控，server的内存，一直在涨，一直到溢出。

我想不明白，为什么更新的是同一缓存的同一id,还会这样？

2 并100的时候，我的was不会出错500错误，设为200，就会挂住。请问有什么参数调整吗？concurrency?

请问是否可以免费进行商用，我用的是open-source.

主要集群ehcache

马老师，您好。

我作的是ehcache的集群，我在tomcat下，好用。

我用的是was6 企业版的，用的是最新的terracotta...

在启动时，出那个异常信息。望老师帮助。

另:第三个问题，已经解决了，是因为，在linux下，我用了was的jdk,改成sun的就好用了。

1 tc-config.xml

<tc-config xmlns="http://www.terracotta.org/config">
<clients xmlns:tc="http://www.terracotta.org/config" xmlns:con="http://www.terracotta.org/config" xmlns="">
<logs>%(user.home)/terracotta/client-logs</logs>
<modules>
<module name="terracotta-toolkit-1.1" group-id="org.terracotta.toolkit"/>
</modules>
</clients>
<servers xmlns:tc="http://www.terracotta.org/config" xmlns:con="http://www.terracotta.org/config" xmlns="">
<server host="192.168.1.2" name="192.168.1.2:9510" bind="0.0.0.0">
<data>%(user.home)/terracotta/server-data</data>
<logs>%(user.home)/terracotta/server-logs</logs>
<statistics>%(user.home)/terracotta/server-statistics</statistics>
</server>
<mirror-groups>
<mirror-group>
<ha>
<mode>networked-active-passive</mode>
<networked-active-passive>
<election-time>5</election-time>
</networked-active-passive>
</ha>
<members>
<member>192.168.1.2:9510</member>
</members>
</mirror-group>
</mirror-groups>
<update-check>
<enabled>true</enabled>
<period-days>7</period-days>
</update-check>
</servers>
</tc-config>

2 was起动，出下面的信息

[11-1-14 9:54:52:335 CST] 00000015 ClientImpl W Thread has no 'inheritedAccessControlContext' field - L1 shutdown will fail to
unload classes - see: http://code.google.com/p/guava-libraries/issues/detail?id=92
java.lang.NoSuchFieldException: inheritedAccessControlContext
at java.lang.Class.getDeclaredFieldImpl(Native Method)
at java.lang.Class.getDeclaredField(Class.java:480)
at org.terracotta.express.ClientImpl.<clinit>(ClientImpl.java:47)
at java.lang.J9VMInternals.initializeImpl(Native Method)
at java.lang.J9VMInternals.initialize(J9VMInternals.java:177)
at org.terracotta.express.ClientFactoryImpl.newClient(ClientFactoryImpl.java:146)
at org.terracotta.express.ClientFactoryImpl.createClient(ClientFactoryImpl.java:140)
at org.terracotta.express.ClientFactoryImpl.getOrCreateClient(ClientFactoryImpl.java:116)
at org.terracotta.express.ClientFactory.getOrCreateClient(ClientFactory.java:28)
at net.sf.ehcache.terracotta.StandaloneTerracottaClusteredInstanceFactory.<init>(StandaloneTerracottaClusteredInstanceFactor
y.java:29)
at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:67)
at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)

3 另外

在Mirror group中，能查看到图标是绿色的，但是，点选这个节点右面的Main中去是，Not connected状态。

也看不到，Ehcache的状态。