iradium, smared,
Are you also seeing the problem happen exactly at the top of the hour?
smared,
Can you show me the output of the command “java -version -server”?
It would not hurt to upgrade to the latest Java and Wowza dev build.
Oh my, I’ve thought it’s something to do with my custom code, this is exactly what is happening to us. After a live broadcasting using a stream.play API then after a couple of hours the server dies. No error on the Logs or something. And I’m still investigating using jconsole and it seem nothing wrong as far as I’m monitoring it. Maybe the admin should look on this because this actually affects lots of your customers.
We have similar issue too, Wowza 3 server on Windows 2008 R2 box (Perpetual License) is running smoothly for a while,then JVM is getting restarted without giving any indications like high resource utilization/access violations etc. No entries in error log or access log. Any one of you noticed similar kind of entries in wrapper log?
ERROR | wrapper | 2011/12/30 18:27:27 | JVM appears hung: Timed out waiting for signal from JVM.
ERROR | wrapper | 2011/12/30 18:27:27 | JVM did not exit on request, terminated
STATUS | wrapper | 2011/12/30 18:27:32 | Launching a JVM…
Not in our case, we are performing load test and turned off the video streaming but we have an HTTPProvider2Base and ModuleBase server extensions. We set the wrapper log to Debug level and ran the test again, it does show one outofmemory exception but the JVM hung 4 minutes after that happened.
DEBUG | wrapperp | 2011/12/31 00:02:33 | read a packet PING : ok
DEBUG | wrapper | 2011/12/31 00:02:33 | Got ping response from JVM
DEBUG | wrapperp | 2011/12/31 00:02:34 | send a packet PING : ping
INFO | jvm 1 | 2011/12/31 00:02:35 | Exception in thread “RMI RenewClean-[10.133.70.12:8084]” java.lang.OutOfMemoryError: PermGen spaceINFO | jvm 1 | 2011/12/31 00:02:38 |
DEBUG | wrapperp | 2011/12/31 00:02:39 | send a packet PING : ping
DEBUG | wrapperp | 2011/12/31 00:02:40 | read a packet PING : ok
DEBUG | wrapper | 2011/12/31 00:02:40 | Got ping response from JVM
DEBUG | wrapperp | 2011/12/31 00:02:43 | send a packet PING : ping
DEBUG | wrapperp | 2011/12/31 00:02:48 | send a packet PING : ping
INFO | jvm 1 | 2011/12/31 00:02:49 | Received a packet PING : ping
INFO | jvm 1 | 2011/12/31 00:02:49 | Send a packet PING : ok
DEBUG | wrapperp | 2011/12/31 00:02:52 | send a packet PING : ping
……………………
INFO | jvm 1 | 2011/12/31 00:06:26 | Received a packet PING : ping
INFO | jvm 1 | 2011/12/31 00:06:26 | Send a packet PING : ok
DEBUG | wrapperp | 2011/12/31 00:06:27 | send a packet PING : ping
DEBUG | wrapperp | 2011/12/31 00:06:31 | send a packet PING : ping
DEBUG | wrapperp | 2011/12/31 00:06:36 | send a packet PING : ping
DEBUG | wrapperp | 2011/12/31 00:06:40 | send a packet PING : ping
DEBUG | wrapperp | 2011/12/31 00:06:45 | send a packet PING : ping
DEBUG | wrapperp | 2011/12/31 00:06:49 | send a packet PING : ping
ERROR | wrapper | 2011/12/31 00:06:54 | JVM appears hung: Timed out waiting for signal from JVM.
ERROR | wrapper | 2011/12/31 00:06:54 | JVM did not exit on request, terminatedDEBUG | wrapperp | 2011/12/31 00:06:55 | server listening on port 32000.
DEBUG | wrapper | 2011/12/31 00:06:55 | Waiting 5 seconds before launching another JVM.
STATUS | wrapper | 2011/12/31 00:06:59 | Launching a JVM…
Sorry forgot to mention, we are running 64Bit JVM - Server Mode and Wowza version is 3.0.2 build866.
java version “1.6.0_26”
Java™ SE Runtime Environment (build 1.6.0_26-b03)
Java HotSpot™ 64-Bit Server VM (build 20.1-b02, mixed mode)
Tough sell in our company to upgrade software in the last minute.
Hi
Check your crontab I suspect that this is a daily process at a specific time that tells the Java process to attempt a “HUP”.
However all that happens at this point is that Java will stop as a “HUP” to Java means shutdown.
You can try this by typing (as root) “crontab -l” and from there determine what the problem is.
Jason
Hi
If you are still struggling you can try the following but I have to point out that this is your risk to take
and that you should have a good idea on the subject before pursuing.
try this as root in the root home directory:
“crontab -l > crontab_current”
“crontab -r”
See if it happens again, If it does its not your crontab.
If it doesn’t happen again, it is a process run by your crontab and will be easier to track down.
To put the crontab back as it was do the following:
“crontab crontab_current”
Jason
We have the issue exactly as above … Help
Hi,
i brought a License last week and since then the issue stated happening.
Check with the url you suppled
Wowza Media Server 3 Monthly Edition 3.0.3 build882
Thanks
Paul
FYI - We have finally resolved this issue. We had our server rebuilt three times. First two times we still experienced the same issues. The third time we had it rebuilt without the virtualization layer and now Wowza stopped hanging. It’s been more than a week and everything is still working as expected. Not sure if it was something within the virtual layer or something else that was different between the two configurations.
We’re having similar issues. What does this mean rebuilt without the virutalization layer? How can we do this?
In our case, Wowza does not report anything in the error or access logs, but just stops streaming. Can’t access port 8086 either. Needs to be restarted then all fine.
