Monday, September 15, 2014

Understanding Out Of Memory Messages

If the application execution time is taking more time, or if the Operating System is performing slow and slow, there may be a case of Out of memory. In other words we can say that memory is being allocated but is not being returned when it is no longer used or needed. At last the application of the System will run out of memory causing application to terminate abnormally.

Out of Memory
Normally Out of memory is thrown when there is insufficient space available to allocate in the Java heap or in area of heap. The Garbage collector cannot make any further space available to allow new objects and we can also think that heap cannot be expanded at this point.

When ever a OOM is thrown a stack trace is also printed. In order to diagnose a Out of Memory condition, it is good to analyze the Stack Trace and what error was generated.

After an OutOfMemoryError condition, the JVM can be in an unknown or unreliable state. Even if the JVM appears to be operating normally, it is recommended that you restart the Java application to ensure a clean environment

OOM: Out of Heap Space
This message says that an Object could not be allocated in the Java Heap. There is couple of cases here. At this Point it is not necessary sure that there is memory leak. This Out of heap Space can be due to wrong configuration where a smaller heap was configured for the application. It is always good to tune the application (load test) before finalizing the heap space for the application.

In other cases when an application is running for long time, the application may hold references to Objects that are unable to garbage collect. In these cases the memory that these objects are being used cannot be reclaimed by the Garbage collector and causes the memory to raise and Out of memory.

The Api’s that are called by the application could also hold object references causing memory leaks. In other cases Finalize methods can also cause Out of memory leaks. When an class has a Finalize method, these  type of objects are not reclaimed during Garabage collection,instead they all are added to the finalize Queue which is a daemon thread that are run after the Garbage collection. If the finalize thread cant keep up to the finalize objects, then it may throw up the Out of Heap Space errors.

Find Out objects available for finialization
Dev:vx1423:djbs002-webapps $ jmap -finalizerinfo 32576
Number of objects pending for finalization: 0

Heap dump is created in the directory where the jmap command is issued.
  • Sun JDK 1.5: jmap -heap:format=b <JAVA_PID>
  • Sun JDK 1.5 64-bit: jmap -J-d64 -heap:format=b <JAVA_PID>
  • JRocket : jrcmd <PID> hprofdump filename=/tmp/abc.hprof
The heap dump files will be large sized.

OOM: Out of Perm Space
PermSpace is a area which does not belong to the heap Space. This is a different area configured with -XX:MaxPermSize option. The permSpace area is an area that stores class and method objects. So when ever an object of a class is created in Heap Space , the corresponding Class and its meta data is loaded into the Perm Space. Out of Perm Space arrives when there is no sufficient space available in order to allocate class metadata information for a newly created Object in Heap.

This may be due to some causes like,
  • The application has many classes (e.g. many JSPs).
  • There are many static member variables in classes.
  • There is heavy use of String.intern().
  • There is a classloader leak. If an object outside the classloader holds a reference to an object loaded by the classloader, the classloader cannot be collected.
  • There are many isolated classloaders.
  • The CMS low pause collector by default does not collect the permanent generation space.
  • Singleton patterns where a class has a static reference to the class itself, making the class unloadable.
Perm Space Information can be obtained using
jmap -permstat <PID> >> /tmp/dump

Once we open the file we can see some thing like this

44496 intern Strings occupying 4507848 bytes.
class_loader    classes bytes   parent_loader   alive?  Type

0xd5764d68  46  257496  0xd57adba0  dead     ContextTypeMatchClassLoader
0xd47bcbe8   8   16720    0xd4618170  dead    ContextOverridingClassLoader@0xc52b4ef0
0xe5060348  3    6944     0xe4cad700     live     ContextOverridingClassr@0xd5d823a8

total = 321 10000   41331344        N/A     alive=114, dead=207     N/A

For each class loader object, the following details are printed:
1. The address of the class loader object – at the snapshot when the utility was run.
2. The number of classes loaded (defined by this loader with the method
3. The approximate number of bytes consumed by meta-data for all classes loaded by this class
4. The address of the parent class loader (if any).
5. A “live” or “dead” indication – indicates whether the loader object will be garbage collected in the future.
6.     The class name of this class loader.

OOM: GC OverHead limit exceeded 
 “GC overhead limit exceeded” is generally thrown by the Garbage Collectors that we use for the JVM. This is generally thrown by serial or parallel collectors.

The issue comes when more amount of time is spent in doing the Garabage Collection and very less than 2% of the heap is recovered.

This may be due to applications that are running very long time or may be due to threads that are struck. Due to these sort of threads, the objects that are loaded are not reclaimed and are hold-ed by these struck threads for a long time

The serial or parallel collectors throw this exception and this feature is designed to prevent applications from running for an extended period of time while making little or no progress because the heap is too small. If necessary, this feature can be disabled by adding the option -XX:-UseGCOverheadLimit to the command line.

If the new generation size is explicitly defined with JVM options, decrease the size or remove the relevant JVM options entirely to un-constrain the JVM and provide more space in the old generation for long lived objects.

If there is unintended object retention , we need to check code for changes If the retention looks normal, and it is a load issue, the heap size would need to be increased.

Note : The New Generation Size is specified by -XX:NewSize=n

Set this value to a multiple of 1024 that is greater than 1MB. As a general rule, set -XX:NewSize to be one-fourth the size of the maximum heap size. Increase the value of this option for larger numbers of short-lived objects.

Be sure to increase the New generation as you increase the number of processors.

To find what values were given to the JVM use jmap like

#$ Jmap 9977

Heap Configuration:
MinHeapFreeRatio = 40
MaxHeapFreeRatio = 70
MaxHeapSize = 268435456 (256.0MB)
NewSize = 1048576 (1.0MB)
MaxNewSize = 4294901760 (4095.9375MB)
OldSize = 4194304 (4.0MB)
NewRatio = 8
SurvivorRatio = 8
PermSize = 16777216 (16.0MB)
MaxPermSize = 134217728 (128.0MB)

OOM: Requested array size exceeds VM limit
This message says that application tried to allocate an array size in heap that cannot be created. In most cases these issues arise due to the wrong memory configuration or a bug in application code part where it creates a array with that size.

OOM: requested 32756 bytes for ChunkPool::allocate. Out of swap space?
Possible causes:
- not enough swap space left, or
- kernel parameter MAXDSIZ is very small.
This message says that the JVM cannot allocate native memory for the operation.  The message indicates the size (in bytes - 32765) of the request that failed and the reason for the memory request. In most cases when this type of error comes , the JVM will invoke the fatal error handling mechanism which will generate a fatal error log file which contains information like threads, heap and process information.

This can be considered as a problem with the Operating System which cannot allocate enough native swap space which might be being used heavily by other process or might be insufficient swap space. During these cases we need to depend on the Operating System tools to analyze the issues.

OOM: Unable to create new native thread or StackOverflowException or Attempt to ungaurd stack red zone failed
A StackOverflowError occurs when the size if the memory that the stack of a Java program require is greater than the memory size set by the Java Runtime Environment.

These occur mainly due to deeply-nested applications or through infinite loops created in error by the programmer in the application. Graphics intensive programs commonly require larger Java stacks than the default value created by the Runtime Environment

Variables stored in stacks are only visible to the owner Thread, while objects created in heap are visible to all thread. In other words stack memory is kind of private memory of Java Threads, while heap memory is shared among all threads.

Most cases we can use the –Xss jvm Falg and lower the amount of stack Size that each Threads gets

When I see an OutOfMemoryError, does that mean that the Java program will exit?

Not always. Java programs can catch the exception thrown when OutOfMemory occurs, and (possibly after freeing up some of the allocated objects) continue to run.

No comments :

Post a Comment