tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From André Warnier ...@ice-sa.com>
Subject Re: Please help diagnosing a random production Tomcat 7.0.53 Internal Server Error!
Date Tue, 15 Apr 2014 21:35:11 GMT
Ian Long wrote:
> On April 15, 2014 at 4:58:28 PM, André Warnier (aw@ice-sa.com(mailto:aw@ice-sa.com))
wrote:
> 
>> Ian,
>>  
>> On this list, it is kind of frowned-upon to "top post". It is preferred when people
answer
>> a question, below the question. Keeps things more logical in the reading sequence,
and
>> avoids having to scroll down to guess what you are responding to.
>>  
>> Ian Long wrote:
>>> Yes, I checked both the tomcat log (I’ve configured tomcat to use log4j) as
well as my application logs.
>>>
>>> Yes, 20 httpd prefork processes.
>>>
>>> I don’t think it’s memory related, I have an 8GB heap and tomcat averages
5GB usage and peeks around 6.5 before garbage collection kicks in.
>>>
>>  
>> Of course we do not know (yet) either what the cause of your problem is.
>> But we know that Tomcat would normally write something in its logs, when a server
error
>> 500 happens.
>> So,
>> - either Tomcat and /or your application wrote something to a logfile, and you have
not
>> yet found that logfile
>> - or else Tomcat and/or your application crashed, but did not write anything to the
logs.
>> In that last case, one of the most likely causes for such a behaviour is running
out of
>> memory.
>> Whether you believe that this is possible or not is your opinion.
>> But it is of the nature of software bugs, to be unexpected.
>> If they were expected, they would have been corrected already.
>>  
> 
> Ok, thanks, didn’t know about the top post issue.
> 
> I have tomcat configured to log via log4j, and then there is my application log, those
are the only two logs, and neither contains anything.
> 
> It’s not about believing, I have monitoring software that gives me precise information
about memory use and there is no indication of a problem there.
> 

Would that monitoring software detect a very short occasional spike in the usage of 
memory, just before the thread running that application is blown out of the water and the

memory usage returns to normal ?
Or is it something that updates its data on a 5-second interval and it just always misses

the significant event ?

Honestly, I am just fishing and trying to find a clue (or rather, trying to help you find

a clue). But some problems are just like that. You can only carefully eliminate the 
possible causes one after the other until you're left with one that you cannot eliminate.

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
For additional commands, e-mail: users-help@tomcat.apache.org


Mime
View raw message