forums.smartfoxserver.com

Posted: **23 Nov 2009, 03:35**

Hello,

We are having a nasty problem where Smartfox is being restarted by a sigkill 9. On top of that, it is not restarting correctly. To make things even more problematic it only happens on our live server ( just our luck ) Our host (Peer1 managed) claims there is nothing on the system that could be sending the sigkill 9 signal.

Can you please provide some suggestions as to find out how this happens? Or perhaps fix the problem for some compensation? We are really in a snag here. Log around the problem:

Code: Select all

INFO   | jvm 1    | 2009/11/20 16:08:37 | 16:08:37.294 - [ INFO ] > Ster:2861d5c04b961e53
INFO   | jvm 1    | 2009/11/20 16:08:37 | 16:08:37.294 - [ INFO ] > Ster:c6e766b3f830c5b6
INFO   | jvm 1    | 2009/11/20 16:08:37 | 16:08:37.295 - [ INFO ] > Ster:9d3d3d240a93db03
ERROR  | wrapper  | 2009/11/20 16:09:12 | JVM appears hung: Timed out waiting for signal from JVM.
ERROR  | wrapper  | 2009/11/20 16:09:12 | JVM did not exit on request, terminated
STATUS | wrapper  | 2009/11/20 16:09:12 | JVM received a signal SIGKILL (9).
STATUS | wrapper  | 2009/11/20 16:09:16 | Launching a JVM...
INFO   | jvm 2    | 2009/11/20 16:09:17 | WrapperManager: Initializing...
INFO   | jvm 2    | 2009/11/20 16:09:19 | |::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|
INFO   | jvm 2    | 2009/11/20 16:09:19 | |                                                            |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |              ...:::  SmartFoxServer :::...                 |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |                Multiplayer Socket Server                   |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |                      version 1.6.6                         |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |                           ---                              |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |              (c) 2004 - 2009 gotoAndPlay()                 |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |                  www.smartfoxserver.com                    |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |                    www.gotoandplay.it                      |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |                                                            |
INFO   | jvm 2    | 2009/11/20 16:09:19 | |::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|
INFO   | jvm 2    | 2009/11/20 16:09:19 | 
INFO   | jvm 2    | 2009/11/20 16:09:20 | ::::::::::: { BlueBox INITED } ::::::::::::
INFO   | jvm 2    | 2009/11/20 16:09:20 | :                                         :
INFO   | jvm 2    | 2009/11/20 16:09:20 | : Version 1.0.5 -- (c) 2008 gotoAndPlay() :
INFO   | jvm 2    | 2009/11/20 16:09:20 | :                                         :
INFO   | jvm 2    | 2009/11/20 16:09:20 | :::::::::::::::::::::::::::::::::::::::::::
INFO   | jvm 2    | 2009/11/20 16:09:20 | 16:09:20.583 - [ INFO ] > Starting h2 engine...
INFO   | jvm 2    | 2009/11/20 16:09:20 | error:null
INFO   | jvm 2    | 2009/11/20 16:09:20 | java.lang.NullPointerException
INFO   | jvm 2    | 2009/11/20 16:09:20 | 	at it.gotoandplay.smartfoxserver.extensions.ExtensionHelper.<init>(ExtensionHelper.java:79)
INFO   | jvm 2    | 2009/11/20 16:09:20 | 	at it.gotoandplay.smartfoxserver.extensions.ExtensionHelper.instance(ExtensionHelper.java:90)
INFO   | jvm 2    | 2009/11/20 16:09:20 | 	at com.veltive.eccky.DBManager.<init>(DBManager.java:51)
INFO   | jvm 2    | 2009/11/20 16:09:20 | 	at com.veltive.eccky.DBManager.<clinit>(DBManager.java:41)

Posted: **23 Nov 2009, 09:55**

Hi,
I highly suspect that they do have a daemon killing applications that they "don't like". If I recall correctly RedHat has a system like this that terminates applications that take too many resources etc...
Also there is no other explanation, as a SIG_KILL is not auto-generated.

On top of that, it is not restarting correctly.

Well, after the SIG_KILL the application shouldn't restart at all.

In order to see if there is any compatibility problems with the native wrapper (that launches the JVM) I would suggest to temporarily run the server without it, by using the start.sh script instead.

Btw, what Linux OS is this?

Posted: **24 Nov 2009, 01:26**

Hi,

This is indeed Red Hat, however our host claims there is nothing that should be killing our process. I can't seem to find anything really about red hat killing of programs. Do you have some references I can check, or perhaps some commands I can use to check if such a function is in place?

Posted: **24 Nov 2009, 07:52**

I am no RedHat expert so I am not able to point you to right thing here, sorry. Also I would presume that if the hosting have double checked they should be right... at least I hope.
My recommendation is to perform the test I suggested earlier. Try launching the server directly, via start.sh without any native wrapper and see what happens. If the JVM gets quitted after some time there's definitely something weird going on.

Let us know how it goes.

forums.smartfoxserver.com

[licensed] Sigkill 9 problem

[licensed] Sigkill 9 problem