NPE due to incorrect condition in java

[PhantomYdn]

Hello,

Please check the following 2 lines of code

orientdb/core/src/main/java/com/orientechnologies/orient/core/storage/impl/local/statistic/OSessionStoragePerformanceStatistic.java

Lines 1197 to 1198 in add35f9

	if (writCacheCountersHolder != null)
	walCountersHolder.clean();

As you can see: you can get easily NPE if walCountersHolder==null (our case).

Full exception:

Uncaught exception in thread 'OrientDB Write Cache Flush Task (Orienteer)'
java.lang.NullPointerException
        at com.orientechnologies.orient.core.storage.impl.local.statistic.OSessionStoragePerformanceStatistic.makeSnapshotIfNeeded(OSessionStoragePerformanceStatistic.java:1236)
        at com.orientechnologies.orient.core.storage.impl.local.statistic.OSessionStoragePerformanceStatistic.stopWriteCacheFlushTimer(OSessionStoragePerformanceStatistic.java:716)
        at com.orientechnologies.orient.core.storage.cache.local.OWOWCache$PeriodicFlushTask.run(OWOWCache.java:1931)
        at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
        at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
        at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
        at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
        at java.lang.Thread.run(Thread.java:748)

[PhantomYdn]

Btw, NPE is also possible if monitoring was not started for some reason (exception and etc), but stopMonitoring has been invoked due to the following construction:

orientdb/core/src/main/java/com/orientechnologies/orient/core/storage/impl/local/statistic/OPerformanceStatisticManager.java

Lines 291 to 301 in 3d01004

	public void stopMonitoring() {
	switchLock.acquireWriteLock();
	try {
	enabled = false;
	final PerformanceCountersHolder countersHolder = ComponentType.GENERAL.newCountersHolder();
	final Map<String, PerformanceCountersHolder> componentCountersHolder = new HashMap<>();
	WritCacheCountersHolder writCacheCountersHolder = deadThreadsStatistic.writCacheCountersHolder;
	StorageCountersHolder storageCountersHolder = deadThreadsStatistic.storageCountersHolder;
	WALCountersHolder walCountersHolder = deadThreadsStatistic.walCountersHolder;

I think it will be safer if stopMonitoring() will exit immediately if monitoring was not enabled.

[PhantomYdn]

I'm sorry, but it seems that OPerformanceStatisticManager was not fully tested, because due to mentioned NPE - it's not possible to get OWOWCache$PeriodicFlushTask running - it stops execution after first Exception (btw - why? Critical things, imho, should be retryable) - so if performance monitoring was started - WOW Cache stops immediately flushing. Am I missing something?

[andrii0lomakin]

Fixed

[PhantomYdn]

Nice fix;) Thank you!