ZanderXu opened a new pull request, #4597:
URL: https://github.com/apache/hadoop/pull/4597

   RouterRpcFairnessPolicyController supports configurable permit acquire 
timeout. Hardcode 1s is very long, and it has causes an incident in our prod 
environment when one nameserivce is busy.
   
   And the optimal timeout maybe should be less than p50(avgTime).
   
   And all handlers in RBF is waiting to acquire the permit of the busy ns. 
   ```
   "IPC Server handler 12 on default port 8888" #2370 daemon prio=5 os_prio=0 
tid=? nid=?  waiting on condition [?]
      java.lang.Thread.State: TIMED_WAITING (parking)
        at sun.misc.Unsafe.park(Native Method)
        - parking to wait for  <?> (a 
java.util.concurrent.Semaphore$NonfairSync)
        at 
java.util.concurrent.locks.LockSupport.parkNanos(LockSupport.java:215)
        at 
java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedNanos(AbstractQueuedSynchronizer.java:1037)
        at 
java.util.concurrent.locks.AbstractQueuedSynchronizer.tryAcquireSharedNanos(AbstractQueuedSynchronizer.java:1328)
        at java.util.concurrent.Semaphore.tryAcquire(Semaphore.java:409)
        at 
org.apache.hadoop.hdfs.server.federation.fairness.AbstractRouterRpcFairnessPolicyController.acquirePermit(AbstractRouterRpcFairnessPolicyController.java:56)
        at 
org.apache.hadoop.hdfs.server.federation.fairness.DynamicRouterRpcFairnessPolicyController.acquirePermit(DynamicRouterRpcFairnessPolicyController.java:123)
        at 
org.apache.hadoop.hdfs.server.federation.router.RouterRpcClient.acquirePermit(RouterRpcClient.java:1500)
   ```
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to