Welcome!

Java IoT Authors: Pat Romanski, Yeshim Deniz, Liz McMillan, Zakia Bouachraoui, Elizabeth White

Related Topics: Java IoT, Weblogic

Java IoT: Article

Java Thread Dumps

Analyzing Java Thread Dumps

Software maintenance is a tedious and challenging job. As long as the software functions as expected, it looks great. Imagine the situation, your pager keeps buzzing in the midnight (Not a happy feeling, right?).

Any software system, no matter how well it has been built and quality tested might get into run-time performance issues. The reasons can be within the functional boundary or can be from the external environment. Software systems are built on certain assumptions and pre-conceived notions. However, when they go live, the assumptions may turn out to be false causing the system to malfunction.

In enterprise J2EE systems which generally have very large user base and involve interactions between heterogeneous systems, one of the general runtime issues reported is the system slowdown or system "hang". In such situations, the general trouble shooting pattern would be to analyze the Java thread dumps to isolate the threads which are causing the slow-down or hang. This article discusses the Java stack traces; anatomy of Java Threads and how to read thread dumps in general.

Exceptions and Stack Traces
All of us encounter and have encountered exceptions during our learning/development phases. Exception is the way Java reports a runtime error. Exceptions have two parts. Message and the stack trace. Message tells you what has gone wrong. Stack Traces provide a complete call flow through all the classes involved top-down as part of the runtime error.

The following is an example stack trace for an ArrayIndexOutOfBoundsException

Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 4

at Test.run(Test.java:13)

at Test.<init>(Test.java:5)

at Test.main(Test.java:20)

In the above exception, the first line "Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 4" tells you that the JVM encountered the reported exception when it tried to access the value at array index 4. The java thread which encountered the exception is "main" thread.

Next let's read through the stack trace. The rule of thumb for exceptions is to read from the top line (message line) to understand what the exception is and keep going down to understand the call flow. In the above example, the call flow started at Test.java line number 20 (main() method) and then it called the constructor of the Test class. Constructors are represented with <init> in the stack traces. Then it moved on to the run() method in the Test class and encountered the reported exception at line number 13.

From the above stack trace, we can conclude that in Test.java, an attempt is made to read past the size of the Array.

The Java Thread Dump
Java Thread dump can be defined as a snapshot of all threads running inside a Java Virtual Machine (JVM) at a given point of time. A thread dump might contain a single thread or multiple threads. In multi threaded environments like J2EE application servers, there will be many threads and Thread Groups. Each of these threads will have its own call stack and will be performing a separate function at a given time. Thread Dump will provide the stack traces of all the JVM threads and much more information about a particular thread.

The Java VM Process and Java Threads
The Java Virtual Machine or JVM is an operating system level process. Java Threads are child processes or Light-Weight Processes (in Solaris terms) of the JVM process.

Generating Java Thread Dumps
Thread Dump is generated by sending a SIGQUIT signal to the JVM process. There are different ways of sending this signal to the process.

In Unix, use "kill -3 <pid>" where pid is the Process ID of the JVM.

In Windows, press CTRL+BREAK on the window where the JVM is running

Java Thread State
Each Java Thread can be in one of the four states during its lifetime.

Runnable - The thread is either running or ready to run when it gets its CPU turn. JRockit thread dumps refer to this state as ACTIVE.

Waiting on Monitor - The thread is either sleeping or waiting on an object for said period of time or waiting to be notified by another thread. This will happen when any of the sleep() methods on Thread object or wait() method on an Object is called.

For example, in WebLogc server the idle execute threads are in this condition and they wait till a socket reader thread notify them of some new work to be done. The stack trace will look like this:

"ExecuteThread: '2' for queue: 'weblogic.admin.RMI'" daemon prio=5 tid=0x1752F040 nid=0x180c in Object.wait() [1887f000..1887fd8c]

at java.lang.Object.wait(Native Method) waiting on <04134D98> (a weblogic.kernel.ExecuteThread)

at java.lang.Object.wait(Object.java:426)

at weblogic.kernel.ExecuteThread.waitForRequest(ExecuteThread.java:126)

locked <04134D98> (a weblogic.kernel.ExecuteThread)

at weblogic.kernel.ExecuteThread.run(ExecuteThread.java:145)

Some other versions of the JVM call this state as CW, Object.wait() (as above). JRockit refer to this as state as WAITING.

Waiting for Monitor Entry - The thread is waiting to get the lock for an object (some other thread may be holding the lock). This happens if two or more threads try to execute synchronized code. Note that the lock is always for an object and not for individual methods.

Sample stack trace of a thread in this condition:

"ExecuteThread: '24' for queue: 'DisplayExecuteQueue'" daemon prio=5 tid=0x5541b0 nid=0x3b waiting for monitor entry [49b7f000..49b7fc24]

at weblogic.cluster.replication.ReplicationManager.createSecondary (ReplicationManager.java:908)
- waiting to lock <6c4b9130> (a java.lang.Object)
at weblogic.cluster.replication.ReplicationManager.updateSecondary (ReplicationManager.java:715)
at weblogic.servlet.internal.session.ReplicatedSessionData.syncSession (ReplicatedSessonData.java:459)
- locked <6c408700> (a weblogic.servlet.internal.session.ReplicatedSessionData)
at weblogic.servlet.internal.session.ReplicatedSessionContext.sync (ReplicatedSessionContext.java:134)
- locked <6c408700> (aweblogic.servlet.internal.session.ReplicatedSessionData)
at weblogic.servlet.internal.ServletRequestImpl.syncSession (ServletRequestImpl.java:2418)
at weblogic.servlet.internal.WebAppServletContext.invokeServlet (WebAppServletContext.java:3137)
at weblogic.servlet.internal.ServletRequestImpl.execute (ServletRequestImpl.java:2544)
at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:153)
at weblogic.kernel.ExecuteThread.run(ExecuteThread.java:134)

In the above stacktrace, you can see that this thread holds the lock an object (6c408700) and waiting to lock another object (6c4b9130)

Some other JVMs may not give the object Ids with lock information in the stack trace. The same state may also be called as ‘MW'. JRockit refers to this state as LOCKED.

Anatomy of a Java Thread
To be able to read/analyze thread dump, it is important to understand various parts of a Thread dump. Let's take a simple Thread Stack example and read various parts of it.

"ExecuteThread: '1' " daemon prio=5 tid=0x628330 nid=0xf runnable [0xe4881000..0xe48819e0]

at com.vantive.vanjavi.VanJavi.VanCreateForm(Native Method)

at com.vantive.vanjavi.VanMain.open(VanMain.java:53)

at jsp_servlet._so.__newServiceOrder.printSOSection( __newServiceOrder.java:3547)

at jsp_servlet._so.__newServiceOrder._jspService (__newServiceOrder.java:5652)

at weblogic.servlet.jsp.JspBase.service(JspBase.java:27)

at weblogic.servlet.internal.ServletStubImpl.invokeServlet (ServletStubImpl.java:265)

at weblogic.servlet.internal.ServletStubImpl.invokeServlet (ServletStubImpl.java:200)

at weblogic.servlet.internal.WebAppServletContext.invokeServlet (WebAppServletContext.java:2495)

at weblogic.servlet.internal.ServletRequestImpl.execute (ServletRequestImpl.java:2204)

at weblogic.kernel.ExecuteThread.execute (ExecuteThread.java:139)

at weblogic.kernel.ExecuteThread.run(ExecuteThread.java:120)

In the above Thread Dump, the interesting part to is the first line. The rest of the stuff is nothing more than a general stack trace. Lets analyze the first line here

Execute Thread : 1 This indicates the name of the thread.

daemon Indicates that this Thread is a daemon thread

prio=5 The priority of this thread (default is 5)

tid The Java Thread Id (Unique identifier for this thread in the running JVM instance).

nid Native Identifier of the thread. Indicates LWP id in Solaris. Identifier for this process

at the OS level.

runnable Indicate the Thread state (See above)

[x..y] Indicates the range of addresses in the heap where this thread is executing.

The remaining part of the thread dump show the call flow. In this example, the thread (Execute Thread 1) is a OS daemon which is currently executing a native method vanCreateForm().

Putting Thread Dumps to Use
In this section I describe few use cases in which Thread dumps are very useful.

High CPU consumption
Diagnosis

The application seems to consume almost 100% CPU and the throughput has come down drastically. Starts to perform very poorly with high load on the CPU.

Thread Dump
All threads in the dump generally show one or more threads struck doing the same operation through all the thread dumps.

Solution

  • Take a series of thread dumps (around 5-7) for one particular call flow (a web form submit for example) before the call is completed.
  • Looking for "runnable" threads in the Thread Dump. If each such thread seems to move (the method call in each thread dump is different), then the thread is progressing and probably not a culprit. If the thread is executing the same method (same line number) through all the thread dumps, then it is almost for sure the culprit. Go to the code and do the code level analysis. You are almost certain to catch an issue there.

Low CPU consumption and Poor Response time
Diagnosis

This happens in a high I/O bound systems under high load. The CPU consumption is low with only few threads consuming reasonable CPU chunk. However the application response time will be very slow.

Thread Dump
Some or all runnable threads seem to struck performing an I/O operation like a File read/write or a database operation.

Solution
Understand the I/O operation profile of your application. Use caches where applicable to reduce DB interactions.

Application/Server Hang
Diagnosis

An application or an server JVM hosting the application will hang (become unresponsive).

Thread Dump

  • All runnable threads are struck doing the same operation through all the thread dumps taken. The server runs out of any more threads as all the runnable threads "never" complete their operation.
  • There might be many threads waiting for monitor entry. This can happen if a "runnable" thread is holding a lock on an object and never returns it while others are waiting for the same lock.

Solution

  • Check for dead lock. JVM generally detects a deadlock if it is simple enough (Thread A waiting for Thread B and vice versa). However, you need to understand the lock profile at the given moment to see if there is any complex deadlock situation involved.
  • Revisit the synchronized methods/block in the code. Reduce the size of synchronization area to the extent possible.
  • One issue could be too long a timeout while accessing a remote resource/component. Implement a reasonable timeout on the remote object clients so that the throw appropriate exception when a remote system does not respond in a reasonable amount of time.
  • If all the threads are waiting for a resource (like an EJB/DB connection), consider increasing the object pool size for those resources.

Tools
There are both commercial and open source tools available for thread dump analysis. One such tool is Samurai. It is a light weight open source tool which runs as a Java Web Start application as well as from your command prompt. For more information and documentation on Samurai, visit

http://yusuke.homeip.net/samurai/en/index.html

Conclusion
Maintaining J2EE enterprise application in a production environment is a tough job. As the business dynamics change, J2EE application environments change which might cause runtime instability in the production application. One of the primary factors that effect an running J2EE application is high load. While most systems are designed to be scalable, environmental limitations might cause them to become non-responsive.

Java Thread Dumps is an excellent mechanism to identify, diagnose, detect and resolve typical production issues. While application profiling and other mechanisms do exist, analyzing Java Thread dumps will provide us an clear and early understanding of popular production level issues there by saving time and helping us provide better user experience with the production applications.

More Stories By Shankar Itchapurapu

Shankar Itchapurapu is a software engineer at Oracle in India. He holds a Master's degree in Computer Applications. You can e-mail Shankar at [email protected]

IoT & Smart Cities Stories
Early Bird Registration Discount Expires on August 31, 2018 Conference Registration Link ▸ HERE. Pick from all 200 sessions in all 10 tracks, plus 22 Keynotes & General Sessions! Lunch is served two days. EXPIRES AUGUST 31, 2018. Ticket prices: ($1,295-Aug 31) ($1,495-Oct 31) ($1,995-Nov 12) ($2,500-Walk-in)
Andrew Keys is Co-Founder of ConsenSys Enterprise. He comes to ConsenSys Enterprise with capital markets, technology and entrepreneurial experience. Previously, he worked for UBS investment bank in equities analysis. Later, he was responsible for the creation and distribution of life settlement products to hedge funds and investment banks. After, he co-founded a revenue cycle management company where he learned about Bitcoin and eventually Ethereal. Andrew's role at ConsenSys Enterprise is a mul...
Nicolas Fierro is CEO of MIMIR Blockchain Solutions. He is a programmer, technologist, and operations dev who has worked with Ethereum and blockchain since 2014. His knowledge in blockchain dates to when he performed dev ops services to the Ethereum Foundation as one the privileged few developers to work with the original core team in Switzerland.
René Bostic is the Technical VP of the IBM Cloud Unit in North America. Enjoying her career with IBM during the modern millennial technological era, she is an expert in cloud computing, DevOps and emerging cloud technologies such as Blockchain. Her strengths and core competencies include a proven record of accomplishments in consensus building at all levels to assess, plan, and implement enterprise and cloud computing solutions. René is a member of the Society of Women Engineers (SWE) and a m...
Digital Transformation and Disruption, Amazon Style - What You Can Learn. Chris Kocher is a co-founder of Grey Heron, a management and strategic marketing consulting firm. He has 25+ years in both strategic and hands-on operating experience helping executives and investors build revenues and shareholder value. He has consulted with over 130 companies on innovating with new business models, product strategies and monetization. Chris has held management positions at HP and Symantec in addition to ...
The challenges of aggregating data from consumer-oriented devices, such as wearable technologies and smart thermostats, are fairly well-understood. However, there are a new set of challenges for IoT devices that generate megabytes or gigabytes of data per second. Certainly, the infrastructure will have to change, as those volumes of data will likely overwhelm the available bandwidth for aggregating the data into a central repository. Ochandarena discusses a whole new way to think about your next...
CloudEXPO | DevOpsSUMMIT | DXWorldEXPO are the world's most influential, independent events where Cloud Computing was coined and where technology buyers and vendors meet to experience and discuss the big picture of Digital Transformation and all of the strategies, tactics, and tools they need to realize their goals. Sponsors of DXWorldEXPO | CloudEXPO benefit from unmatched branding, profile building and lead generation opportunities.
Dynatrace is an application performance management software company with products for the information technology departments and digital business owners of medium and large businesses. Building the Future of Monitoring with Artificial Intelligence. Today we can collect lots and lots of performance data. We build beautiful dashboards and even have fancy query languages to access and transform the data. Still performance data is a secret language only a couple of people understand. The more busine...
All in Mobile is a place where we continually maximize their impact by fostering understanding, empathy, insights, creativity and joy. They believe that a truly useful and desirable mobile app doesn't need the brightest idea or the most advanced technology. A great product begins with understanding people. It's easy to think that customers will love your app, but can you justify it? They make sure your final app is something that users truly want and need. The only way to do this is by ...
DXWorldEXPO LLC announced today that Big Data Federation to Exhibit at the 22nd International CloudEXPO, colocated with DevOpsSUMMIT and DXWorldEXPO, November 12-13, 2018 in New York City. Big Data Federation, Inc. develops and applies artificial intelligence to predict financial and economic events that matter. The company uncovers patterns and precise drivers of performance and outcomes with the aid of machine-learning algorithms, big data, and fundamental analysis. Their products are deployed...