Hi,
We have a test 6.5.2 system that is failing to start (running on Windows, so currently rebooting the server in case it is a file locking problem or something.) But it is failing within a second or two of launching (either via Windows Services or on the command line) without touching any files, even the console.log. That seems too early to be a database connection problem, but my question is: roughly how soon does Rhythmyx start trying to connect to the database after it is started?
Thanks,
Andrew.
Looking at a couple of out of the box implementations I have with remote databases, Rx starts the binding process after only a couple seconds. Check the <rxroot>/AppServer/server/rx/log/server.log for errors.
The server.log would be easiest to read through if your remove the current server.log while the server is down, then try to start the server. When the server restarts, a new server.log will be created with only the data from the current restart will be included.
Thanks for the info. Our servers appear to be running fine, and are connected to the database, but once stopped they cannot be restarted. Possibly they cannot reconnect to the database, although we can via other clients. The failure to start the Rhythmyx service happens before it writes anything to any log file. Literally nothing in the filetree has been modified since the moment the service was last stopped. I’ve already spoken with UK technical support, and it doesn’t seem to be a common problem they have a fix for. It may well be a problem with our operating system, or networks, or database, but any ideas for diagnosis would be appreciated. We have just launched a new web site and, while the production Rhythmyx server is still up, we have to hope it doesn’t need to be stopped, because we probably won’t be able to start again.
Do you have a RhythmyxServer.ja file in the Rhythmyx root? I have found that sometimes if the memory in that file is too high for the server then it won’t start. You can also try to start the server using these additional command line options for more information:
RhythmyxServer.exe -is:log output.log
Even rebooting the box didn’t help anything? Have you tried starting the server by just double clicking the RhythmyxServer.exe?
When was the last time you restarted the Rx server before this behavior started occurring? It’s possible that changes you’ve made since the last successful restart didn’t take effect until this most recent restart. Maybe something didn’t go well. What changes occurred in your system between then and now?
You also might want to try setting up your most recent backup of the environment on the same machine to see if that works. If it does, then something in your recent changes didn’t take well. If it still doesn’t start up, that would indicate that it might be an issue with the box itself (or the backup is messed up as well). In that case, try setting up the backup on a different machine.
[QUOTE=slyons;10698]Do you have a RhythmyxServer.ja file in the Rhythmyx root? I have found that sometimes if the memory in that file is too high for the server then it won’t start. You can also try to start the server using these additional command line options for more information:
RhythmyxServer.exe -is:log output.log[/QUOTE]
Deleting the RhythmyxServer.ja file seems to have done the trick. Thank you very much. But we have had that .ja file in there for months, unmodified, and stopped and started our three servers twice weekly all that time without problems. Maybe there is a problem with the memory on those servers, which I’ll check with our IT Support. Running RhythmyxServer.exe (as I had already tried at the suggestion of UK technical support) but with the logging option you suggest, recorded the following before deleting the .ja file…
file type: 0
file id: 0
length: 278
offset: 299090
file name: instructions.txt
file type: 5
file id: 1
length: 1009
offset: 299368
file name: Verify.jar
file type: 2
file id: 2
length: 406
offset: 300377
file name: launch.txt
file type: 1
file id: 3
length: 1024
offset: 300783
file name: jvm
Verifying JVM at D:\Rhythmyx\JRE\bin\java.exe
for JVM: Java 1.5.0_12 by Sun Microsystems
Executing launch script command: “D:\Rhythmyx\JRE\bin\java.exe” -cp “”;“C:\DOCUME~1\xnm89743\LOCALS~1\Temp\LRE76.tmp\Verify.jar”;“” “Verify” java.vendor java.version in working directory D:\Rhythmyx
Output from the launch script execution
StdOut:
1739021872 Sun Microsystems Inc. 1.5.0_12
ErrOut:
Found matching JVM at D:\Rhythmyx\JRE
After VERIFY, the jvm home is D:\Rhythmyx\JRE
Executing launch script command: “D:\Rhythmyx\JRE/bin\java.exe” -cp “”;JRE/lib/tools.jar;AppServer/bin/run.jar;“” -Dprogram.name=RhythmyxServer.exe -Xms128m -Xmx512m -Djava.endorsed.dirs=“AppServer/lib/endorsed” -Djava.library.path=./bin -server -XX:MaxPermSize=128m -Xmx1300m -Xms512m org.jboss.Main --configuration=rx in current working directory
The only difference having deleted the .ja file, which allowed it to start, were obviously the parameters specified by the .ja file. Instead of the above it used what I presume are the defaults…
-Xms128m -Xmx512m
Please could someone rename this thread to something like “What to check if Rhythmyx fails to start after just one second” as I think it contains useful information for future reference. Either that, or get it Rhythmyx to report back with a useful error message such as “insufficient memory” or something. I initially asked about how quickly Rhythmyx attempts to connect to the database, because we had database connection problems earlier in the week.
This problem has been traced to Sophos when they updated their virus checking engine to 7.6.10. We have reverted to 7.6.9 and this fixes the problem. For once, this is nothing to do with Rhythmyx, but I thought I’d post here in case any other customers get the same problem in the future.