I don't really want to get into a pissing contest over this (mainly because
I have some serious pressure and a rather small aperture, resulting in an
unmatchable pressure/volume ratio), so this is the last thing I will say on
the topic that doesn't relate to a solution for you (which, at the moment,
is not possible given the information you've provided about your issue).
"At least 3 years of giving a chance are not "soon" "
You've only been at it for three years? Written history is 'only' a few
thousand years old and 3 years is small by comparison.
"Operating system is only as good as the applications run on it." True of
false? True... in Windows 95 :-) Few years later even Microsoft handles apps
erros successfully."
If this is what you believe in your heart, then I recommend a wholesale
switch to .NET for you. I wish you all the low switching costs in the
world, and commend you for the faith you put in Microsoft (and their error
messages), as this is a gritty display of loyalty in the face of
overwhelming evidence contradicting your assertion.
"I'm so sad that I think of a _completly_ new code. In PHP!"
PHP is great. So is Zoap and .NET. Careful, though, some of the same
problems you would be shunning Java for would also appear in any Internet
development/application development platform. You know the old adage,
however, if you let it go, and it comes back to you, it's yours forever. If
you switch and you aren't very, very stubborn, you may be Java's forever.
"- the apps have _plenty_ of different urls"
There is a correlation between the size of the application and the time it
takes to properly test the application. The larger the application, the
more it takes to vet application quality. What you have said here is that
you don't want to properly test the application because it's too large and
too much work, and that you are happy to continually restart the application
as a solution. Sounds equitable.
" I can't reproduce such Tomcat hangs on development machine."
And with regularity, what steps lead up to Tomcat hanging? This is the
question that you should try to answer with increasing accuracy each time
Tomcat hangs until you find the issue. Once you do, submit it as a bug, and
give back to the OS community that has been supporting your applications for
3 years.
"- Shall I really examin and monitor the production enviroment?"
Only if you don't care about having server downtime for your users.
"Looks like a "finish it yourself" thing"
try
{
Hacker hacker = new Hacker();
if(hacker.canFinishItHimselfByAskingGoodQuestions())
{
hacker.walkProudly();
hacker.goToBank();
hacker.retireWealthy();
hacker.dieHavingLeftAGoodLegacy();
}
else
{
throw new NotReallyAHackerException();
}
}
catch (NotReallyAHackerException hrahe)
{
return new GoToLawOrBusinessSchoolEvent();
}
Bear Down,
JCD
-----Original Message-----
From: news [mailto:[EMAIL PROTECTED] On Behalf Of Tomasz Nowak
Sent: Tuesday, February 28, 2006 11:25 AM
To: users@tomcat.apache.org
Subject: Re: Tomcat 5.5.x crashes almost every single day.
John C. Dale <[EMAIL PROTECTED]> wrote:
>
> Don't give-up on Tomcat so soon.
At least 3 years of giving a chance are not "soon"... ;]
> Sometimes Tomcat is only as good as the applications deployed within
> her.
"Operating system is only as good as the applications run on it." True of
false? True... in Windows 95 :-) Few years later even Microsoft handles apps
erros successfully.
Tomcat is the only server-type software I know that
_any_user_ can crash so easily and the only
that is so helpless about handling their erros.
Why don't it send 500 status to pages in faulty webapp
and contiunes to serve other webapps normally?
> I write my own apps ;o)
I use Apache Foundation apps ;]
> I would stick with it, and cease development of new code until you
> have eliminated the existing installed code
I'm so sad that I think of a _completly_ new code. In PHP!
> Along the way, it would
> be helpful for you to beat up a separate, mirrored staging instance of
> your production environment using JMeter. Reproduce your problem with
> regularity by hitting each URL many times, then examining the results.
This is probably imposibble because:
- the apps have _plenty_ of different urls.
- I can't reproduce such Tomcat hangs on development machine.
- Shall I really examin and monitor the production enviroment?
What about performance then?
Are there any resource monitors built-in into Tomcat?
I've browsed the JMX management manual.. well is it written
for a human being? Looks like a "finish it yourself" thing. :|
--
Tomasz Nowak
Netventure, http://www.netventure.pl/
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
