How much memory is on the server? Your file buffer may be too large and
your disks can not keep up. I think the default is 10% of memory, so if
you can fill that up faster than the disks can write it it will hang
until the buffer is emptied. I doubt you are doing that much file i/o
through a
Caldarale, Charles R wrote:
From: Scott Mueller [mailto:[EMAIL PROTECTED]
Subject: Re: Running out of memory too easily in Tomcat 6.0.16
Charles, where's the proper place to put the heap/permgen
allocation settings in tomcat's scripts?
You don't modify the scripts themselves. Set whatev
Caldarale, Charles R wrote:
From: Enrique Arizón [mailto:[EMAIL PROTECTED]
Subject: Tomcat tool similar to "top"
Time to time one of those apps gets out of control,
but standard OS tools like top just show the java
process running up to 200% of CPU usage.
I'd be suspicious of that tool :
You can't do name-based virtual hosting with SSL. I think I remember
reading that the headers are encrypted and are not decrypted until
after apache would have to make the choice of which virtual host to
send the request to.
I suppose that changing the SSL port would let you use the same IP
addre
If the CPU is maxed out, the thread dump will provide lightweight
process numbers. Using a tool like "prstat -L" on Solaris, you can
figure out which lightweight process (thread) it is, and at least
confirm that it is the GC thread. the output from prstat is indexed
at 1, but the stack dump is in
I doubt this is it, but I would avoid installing tomcat into a
directory containing spaces. It may be looking for the jar file in
"c:\Program"
Mark
On 10/4/07, Jaime Almeida <[EMAIL PROTECTED]> wrote:
> I'm sorry, I didn't mean to upset you.
> But I'm really in trouble with the tomcat server.
>
I think that gmail is balking over the fact that "venus" does not
resolve to 71.248.123.180, and the reverse record of 71.248.123.180
resolves to static-71-248-123-180.bltmmd.east.verizon.net instead of
"venus".
Here's what you want in DNS:
A "MX" (Mail) record to be set to something like "mail.my
I encountered a similar problem, where one servlet had a bug where the
headers would be set _after_ the data had been sent. The result was
that unrelated responses would come back as text/plain. Once I found
the offending code and fixed it, the issue has not surfaced again.
Mark
On 9/24/07, Lar
Once you find them, you might be hard pressed to actually do anything
about it beyond getting in touch with their ISP.
It might be easier to just block them at the firewall or on the server
tomcat runs on with something like iptables.
Mark
On 8/23/07, Lyallex <[EMAIL PROTECTED]> wrote:
> www.who
In addition, there is a minimum page length required for IE to show
your custom error pages. If your error page is less than 512 bytes,
the friendly error page will be shown.
http://support.microsoft.com/kb/294807
Mark
On 7/23/07, David Delbecq <[EMAIL PROTECTED]> wrote:
in IE6 there was a co
I've had it happen in both Internet Explorer and Firefox. Personally
I've only seen it in Firefox, but users had been complaining about the
issue... and they are only using IE.
Mark
On 7/22/07, Pid <[EMAIL PROTECTED]> wrote:
Mark Deneen wrote:
> Since we're all giving
Since we're all giving mod_jk a big hug here I thought I would point
out that I had issues with both mod_proxy_http and mod_proxy_ajp which
were solved by moving (back) to mod_jk.
Intermittently the response to the browser would be sent as plain-text
(the browser just prints out the html) or I wo
It opens up a listening socket on the loopback address. The second
process connects to this socket and sends a message telling tomcat to
shut down.
Mark
On 5/22/07, Johnny Kewl <[EMAIL PROTECTED]> wrote:
=Start Stop Mechanism=
I see that START goes into a wait loop... and it will pop out of th
.
Mark
On 5/16/07, Mark Deneen <[EMAIL PROTECTED]> wrote:
Hi, I have experienced some issues using mod_proxy_ajp and Tomcat
6.0.10 under moderate load. Occasionally I receive an error 503 in my
browser when the webapp when we have around 30 - 40 users connected.
First off, here is what I
Attached is the script I use to start/stop tomcat. The shutdown
method waits a configurable amount of time for the instance to
terminate cleanly and then terminates the process.
Obviously one would need to tweak the script to fit in their
environment. tomcat-prod goes in /etc/init.d and tomcat-
now, I am using mod_proxy_http and have apache connect to the
Tomcat http connector. This works well, but I would like to
understand the issue with the AJP connector. Any ideas? Should I use
mod_jk?
The server is running Centos 4.4 x86_64.
Best Regards,
M
16 matches
Mail list logo
