@Pierre Could you please guide me on how to pin down to a SOA composite when we have a Hogger thread for a long time? I mean what exactly is causing the hogger thread?
Really a very good presentation. But i have a small doubt... How to solve the struck threads?? What is the solution to remove struck threads and make thread health is FINE.
Hi Pierre, very good presentation of thread monitoring. I have a scenario where i don't see hogging threads, execute thread counts are steady but the idle thread count have dropped to zero. This behavior is for almost an hour. During this time, pending user requests are increasing and my weblogic server have received no requests. Any pointers on how i can start debugging would help. Thanks.
Hi Natraj, I would recommend to get a Thread Dump and look at the current requests. There might be a situation of blockage for the incoming requests that requires thread dump analysis to pinpoint the root cause. Also, please look at your JVM and GC utilization at that time. High JVM pause time can easily trigger this type of situation since stop-the-world collections will suspend any Thread processing during this time. Thanks. P-H
+Pierre-Hugues Charbonneau Hi Pierre, Thanks for responding. I did check thread dumps which were generated by dynatrace automatically when the idle thread count dropped down to zero. there were 4 execute weblogic threads in runnable state and they were working on socket read and each one owned a separate monitor lock on java.iobufferedstreamreader. The other execute threads were in waiting state because we did not have enough throughput and they were not found to be waiting on any monitor lock to be acquired.. The heap usage was healthy. minor, major collections, reclaimed memory after collection all looked to be healthy. Consolidating all these, I am now looking into socket connection and their status using netstat logs. Do you have any other pointers which can be checked ?
+Pierre-Hugues Charbonneau Hi Pierre, I am afraid I will not be able to share my screen in TH-cam chat due to security restrictions and currently I am currently out of office till Monday. Will a mail containing screenshots of the thread stack work? If yes can you please share your mail id or any other way to share the screenshot.
Understood, any sensitive data cannot be shared. One suggestion, please ensure that your WLS is using the native Socket Muxer. The problem you describe makes me wonder if your WLS is configured in using Java Socket Muxer which is very not efficient. Look at the stacktrace, if Java Muxer is used, you will see that in the stacktrace. Thanks.
Hi, Is there any way to kill stuck threads. I have web logic server 11g and frequently the server went to warning state due to stuck threads, until the server restart the threads are not being released. Give me the solution how can we avoid this..
A very good presentation, but it would have been good, if you have explained a bit further, how will the thread go backs to ACTIVE status, atleast on what scenarios? offcourse restart of server, will make the thread to ACTIVE, but on what scenarios, does the thread come backs to ACTIVE, without the restart.....?
awsome...Nice Explanation, nice presentation , good pace...loved it
@Pierre Could you please guide me on how to pin down to a SOA composite when we have a Hogger thread for a long time? I mean what exactly is causing the hogger thread?
hey Pierre-Hugues Charbonneau. very good presentation
HI The presentation is very good and understandable and usefull.Thank you soo much.
Excellent presentation Pierre-Hugues,
Richard Langlois P. Eng.
Java Solutions Architect @ Alithya, Montreal.
Thank you Richard, I am glad it helped.
Hi, very good video. It is very informative. Can we the demo program's ? If so we can use it for hands on learning and further research. Thanks
Really a very good presentation. But i have a small doubt... How to solve the struck threads?? What is the solution to remove struck threads and make thread health is FINE.
Excellent .. infect I'm really struggling Hogging and struck thrad problem.
@Pierre-Hugues , I am facing application slowness due to one hogging thread issue.How to resolve hogging thread issue,
Good and explained with real time examples👌👌👌👌👌👌
Please prepare some more videos on jdbc jms topics
Thank you Pierre, If you have time reply to Jagadish Reddy issue.
Hi Pierre, very good presentation of thread monitoring. I have a scenario where i don't see hogging threads, execute thread counts are steady but the idle thread count have dropped to zero. This behavior is for almost an hour. During this time, pending user requests are increasing and my weblogic server have received no requests. Any pointers on how i can start debugging would help.
Thanks.
Hi Natraj,
I would recommend to get a Thread Dump and look at the current requests. There might be a situation of blockage for the incoming requests that requires thread dump analysis to pinpoint the root cause. Also, please look at your JVM and GC utilization at that time. High JVM pause time can easily trigger this type of situation since stop-the-world collections will suspend any Thread processing during this time.
Thanks.
P-H
+Pierre-Hugues Charbonneau Hi Pierre,
Thanks for responding.
I did check thread dumps which were generated by dynatrace automatically when the idle thread count dropped down to zero. there were 4 execute weblogic threads in runnable state and they were working on socket read and each one owned a separate monitor lock on java.iobufferedstreamreader. The other execute threads were in waiting state because we did not have enough throughput and they were not found to be waiting on any monitor lock to be acquired.. The heap usage was healthy. minor, major collections, reclaimed memory after collection all looked to be healthy. Consolidating all these, I am now looking into socket connection and their status using netstat logs. Do you have any other pointers which can be checked ?
Please post the StackTrace of the SoketMuxer Threads in the TH-cam chat windows. I will have a quick look.
Thanks.
+Pierre-Hugues Charbonneau
Hi Pierre,
I am afraid I will not be able to share my screen in TH-cam chat due to security restrictions and currently I am currently out of office till Monday. Will a mail containing screenshots of the thread stack work? If yes can you please share your mail id or any other way to share the screenshot.
Understood, any sensitive data cannot be shared. One suggestion, please ensure that your WLS is using the native Socket Muxer. The problem you describe makes me wonder if your WLS is configured in using Java Socket Muxer which is very not efficient. Look at the stacktrace, if Java Muxer is used, you will see that in the stacktrace.
Thanks.
Hi,
Is there any way to kill stuck threads.
I have web logic server 11g and frequently the server went to warning state due to stuck threads, until the server restart the threads are not being released.
Give me the solution how can we avoid this..
Hello, did you get a solution to your problem? i have something similiar. I don't speak much English but Google Translate helps me
A very good presentation, but it would have been good, if you have explained a bit further, how will the thread go backs to ACTIVE status, atleast on what scenarios? offcourse restart of server, will make the thread to ACTIVE, but on what scenarios, does the thread come backs to ACTIVE, without the restart.....?
Great! Simple! but very Effective. Thanks a Lot:)
Nice demonstration, keep up the good work!
Oracle WebLogic 12c Thread Monitoring...Oracle Weblogic 12c Thread Monitoring
nicely explained...
good explanation
Thank you so much.