Problem Description:
Online service suddenly hung up, can not be called to see if the service logs found Dubbo thread pool all full:
2019-11-22 14:35:26.271 WARN 26516 --- [New I/O server worker #1-2] c.a.d.c.t.support.AbortPolicyWithReport : [DUBBO] Thread pool is EXHAUSTED! Thread Name: DubboServerHandler-192.168.10.26:12350, Pool Size: 200 (active: 200, core: 200, max: 200, largest: 200), Task: 6786 (completed: 6586), Executor status:(isShutdown:false, isTerminated:false, isTerminating:false), in dubbo://192.168.10.26:12350!, dubbo version: 2.6.2, current host: 192.168.10.26
Not much traffic, but the thread has soared, speculation may be where there is a cycle of death or where a deadlock occurred.
First, detect what CPU usage server found in the normal range, basically ruled out where there is a cycle of death.
First find out the process of the service, use jstack
the command dump
thread analysis.
"DubboServerHandler-192.168.10.26:12350-thread-200" #240 daemon prio=5 os_prio=0 tid=0x00007ffa7c141800 nid=0x6c89 waiting on condition [0x00007ffa17c7c000]
java.lang.Thread.State: WAITING (parking)
at sun.misc.Unsafe.park(Native Method)
- parking to wait for <0x00000000e0d24020> (a java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject)
at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
at java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.await(AbstractQueuedSynchronizer.java:2039)
at org.apache.commons.pool2.impl.LinkedBlockingDeque.takeFirst(LinkedBlockingDeque.java:590)
at org.apache.commons.pool2.impl.GenericObjectPool.borrowObject(GenericObjectPool.java:441)
at org.apache.commons.pool2.impl.GenericObjectPool.borrowObject(GenericObjectPool.java:362)
at redis.clients.util.Pool.getResource(Pool.java:49)
at redis.clients.jedis.JedisPool.getResource(JedisPool.java:226)
at (RedisHelper.java:322)
at (RedisHelper.java:106)
at BlockResourceManager.java:54)
- locked <0x00000000e0ec22d0>
You can see the last thread locked <0x00000000e0ec22d0>
, that it holds with the lock, while other states are threads waiting to lock <0x00000000e0ec22d0>
.
It seems that this thread has led to other threads are blocked.
So what is the cause of this thread can not quick-release lock it?
Continue to look at the above information, it should be from the JedisPool
acquisition of JedisClient
, but the client has not been available in the pool end, it is blocked.
The last reason is a method used up a client does not return to the pool, resulting in client resources are exhausted.