[JCS-144] BlockDiskCache hangs on SEVERE: Region [TMS] Failure getting from disk--IOException

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: jcs-2.0-beta-2
• Fix Version/s: jcs-2.0-beta-2
• Component/s: Indexed Disk Cache
• Environment: version 1.7.0_75, vendor Oracle Corporation
• Assignee: Thomas Vandahl
• Reporter: Wiktor N
• Created: 16/Mar/15 22:56
• Updated: 25/Dec/16 18:34
• Resolved: 01/Apr/15 15:10

Description

If I get a failure reading an object from cache, the thread locks, with following stack trace:
"AWT-EventQueue-0" prio=6 tid=0x00000000112aa000 nid=0x1444 waiting on condition [0x0000000013dc9000]
java.lang.Thread.State: WAITING (parking)
at sun.misc.Unsafe.park(Native Method)

• parking to wait for <0x00000007cbf84b00> (a java.util.concurrent.locks.ReentrantReadWriteLock$NonfairSync)
  at java.util.concurrent.locks.LockSupport.park(LockSupport.java:186)
  at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:834)
  at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireQueued(AbstractQueuedSynchronizer.java:867)
  at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquire(AbstractQueuedSynchronizer.java:1197)
  at java.util.concurrent.locks.ReentrantReadWriteLock$WriteLock.lock(ReentrantReadWriteLock.java:945)
  at org.apache.commons.jcs.auxiliary.disk.block.BlockDiskCache.reset(BlockDiskCache.java:643)
  at org.apache.commons.jcs.auxiliary.disk.block.BlockDiskCache.processGet(BlockDiskCache.java:343)
  at org.apache.commons.jcs.auxiliary.AbstractAuxiliaryCacheEventLogging.getWithEventLogging(AbstractAuxiliaryCacheEventLogging.java:109)
  at org.apache.commons.jcs.auxiliary.disk.AbstractDiskCache.doGet(AbstractDiskCache.java:771)
  at org.apache.commons.jcs.auxiliary.disk.AbstractDiskCache.get(AbstractDiskCache.java:279)
  at org.apache.commons.jcs.engine.control.CompositeCache.get(CompositeCache.java:550)
• locked <0x00000007cbf62c00> (a org.apache.commons.jcs.engine.control.CompositeCache)
  at org.apache.commons.jcs.engine.control.CompositeCache.get(CompositeCache.java:455)
  at org.apache.commons.jcs.access.CacheAccess.getCacheElement(CacheAccess.java:125)

Looing at the code in BlockDiskCache I see, that we first acquire read lock in processGet, and then try to acquire write lock in reset(). According to API spec: http://docs.oracle.com/javase/7/docs/api/java/util/concurrent/locks/ReentrantReadWriteLock.html

It is unsupported to upgrade an lock from read to write ("Additionally, a writer can acquire the read lock, but not vice-versa").

Either read lock should be dropped early on exception or different lock mechanism should be used