6000 tickets system slows down and eats memory

Christian_Steger1 · September 21, 2001, 9:47am

hi there,

running my rt 2.0.7 box on a 500mhz box with 512mb ram - redhat 7.1
in the meantime i have already 6000 tickets, some of them contains
attachments.

the problem appears all the time, if i answer an ticket via gui, it needs
in the meantime approx 15 seconds after sending out an answer.
within that time the box eats my whole memory - he start to swap.

other functions within the rt working fine and speedy

sometimes i have seen that some apacheprocesses hanging as child and eats
a part of my cpu as well as mem.

any idea what could be happen here ?

the apache i use is:
Apache/1.3.19 Ben-SSL/1.42 (Unix) mod_perl/1.25

everything is static compiled.

thanks for any comments

christian steger

Jesse_Vincent · September 21, 2001, 3:50pm

A while back, Dave from toad.net posted some indices you could add
which significantly improve performance. Did you add them?
(They are automatically added to new installs these days, too)On Fri, Sep 21, 2001 at 11:47:31AM +0200, Christian Steger wrote:

hi there,

running my rt 2.0.7 box on a 500mhz box with 512mb ram - redhat 7.1
in the meantime i have already 6000 tickets, some of them contains
attachments.

the problem appears all the time, if i answer an ticket via gui, it needs
in the meantime approx 15 seconds after sending out an answer.
within that time the box eats my whole memory - he start to swap.

other functions within the rt working fine and speedy

sometimes i have seen that some apacheprocesses hanging as child and eats
a part of my cpu as well as mem.

any idea what could be happen here ?

the apache i use is:
Apache/1.3.19 Ben-SSL/1.42 (Unix) mod_perl/1.25

everything is static compiled.

thanks for any comments

christian steger

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

http://www.bestpractical.com/products/rt – Trouble Ticketing. Free.

David_C_Troy · September 21, 2001, 3:59pm

The other thing I’ve been doing to keep things running smoothly is running
the following statements nightly via cron – adapt as makes sense for you:

$dbh->do(“OPTIMIZE TABLE Tickets”);
print “Optimized Tickets.\n”;
$dbh->do(“OPTIMIZE TABLE Transactions”);
print “Optimized Transactions.\n”;
$dbh->do(“OPTIMIZE TABLE Attachments”);
print “Optimized Attachments.\n”;
$dbh->do(“OPTIMIZE TABLE Watchers”);
print “Optimized Watchers.\n”;
$dbh->do(“OPTIMIZE TABLE ObjectKeywords”);
print “Optimized ObjectKeywords.\nDone!\n\n”;

This clears out any deleted stuff and defrags the tables too I believe.

I do this at like 3AM and it’s kept things relatively clean.

Dave

David C. Troy [dave@toad.net] 410-544-6193 Sales
ToadNet - Want to go fast? 410-544-1329 FAX
570 Ritchie Highway, Severna Park, MD 21146-2925 www.toad.netOn Fri, 21 Sep 2001, Jesse Vincent wrote:

A while back, Dave from toad.net posted some indices you could add
which significantly improve performance. Did you add them?
(They are automatically added to new installs these days, too)

On Fri, Sep 21, 2001 at 11:47:31AM +0200, Christian Steger wrote:

hi there,

running my rt 2.0.7 box on a 500mhz box with 512mb ram - redhat 7.1
in the meantime i have already 6000 tickets, some of them contains
attachments.

the problem appears all the time, if i answer an ticket via gui, it needs
in the meantime approx 15 seconds after sending out an answer.
within that time the box eats my whole memory - he start to swap.

other functions within the rt working fine and speedy

sometimes i have seen that some apacheprocesses hanging as child and eats
a part of my cpu as well as mem.

any idea what could be happen here ?

the apache i use is:
Apache/1.3.19 Ben-SSL/1.42 (Unix) mod_perl/1.25

everything is static compiled.

thanks for any comments

christian steger

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

–
http://www.bestpractical.com/products/rt – Trouble Ticketing. Free.

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

Christian_Steger1 · September 21, 2001, 4:51pm

hi dave,

thanks for your suggestion, will se whats happen after a while, i have
seen the performance increased now extremly. - putted also as a cron job.

i guess the main problem appears is my large attachmenttable:

38496704 Sep 21 18:31 Attachments.MYD

however, will watch whats going on in the next time.

thanks again to you dave and jesse.

christian

*$dbh->do(“OPTIMIZE TABLE Tickets”);
*print “Optimized Tickets.\n”;
*$dbh->do(“OPTIMIZE TABLE Transactions”);
*print “Optimized Transactions.\n”;
*$dbh->do(“OPTIMIZE TABLE Attachments”);
*print “Optimized Attachments.\n”;
*$dbh->do(“OPTIMIZE TABLE Watchers”);
*print “Optimized Watchers.\n”;
*$dbh->do(“OPTIMIZE TABLE ObjectKeywords”);
*print “Optimized ObjectKeywords.\nDone!\n\n”;

Ayan_R_Kayal · September 21, 2001, 6:00pm

I have my system set to restart Apache every 4 hours, because it slows down
so much. Other people in my organization using RT don’t have the same
problem, but they also don’t have the volume of tickets my setup does
because they’re still in test mode. Incidentally, in test mode, my setup ran
fine without restarting Apache.

O- ~ARK

Jesse_Vincent · September 21, 2001, 6:23pm

Have you looked at cutting down the max # of requests a given apache process
serves before destroying itself and letting a new child take over?

-jOn Fri, Sep 21, 2001 at 02:00:12PM -0400, Ayan R. Kayal wrote:

I have my system set to restart Apache every 4 hours, because it slows down
so much. Other people in my organization using RT don’t have the same
problem, but they also don’t have the volume of tickets my setup does
because they’re still in test mode. Incidentally, in test mode, my setup ran
fine without restarting Apache.

O- ~ARK

-----Original Message-----
From: rt-users-admin@lists.fsck.com
[mailto:rt-users-admin@lists.fsck.com]On Behalf Of Christian Steger
Sent: Friday, September 21, 2001 5:48 AM

hi there,

running my rt 2.0.7 box on a 500mhz box with 512mb ram - redhat 7.1
in the meantime i have already 6000 tickets, some of them contains
attachments.

the problem appears all the time, if i answer an ticket via gui, it needs
in the meantime approx 15 seconds after sending out an answer.
within that time the box eats my whole memory - he start to swap.

other functions within the rt working fine and speedy

sometimes i have seen that some apacheprocesses hanging as child and eats
a part of my cpu as well as mem.

any idea what could be happen here ?

the apache i use is:
Apache/1.3.19 Ben-SSL/1.42 (Unix) mod_perl/1.25

everything is static compiled.

thanks for any comments

christian steger

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

http://www.bestpractical.com/products/rt – Trouble Ticketing. Free.

Ayan_R_Kayal · September 23, 2001, 6:28pm

Jesse,
Thanks for the suggestion. The box appears to be running smoothly now.
Before it had been set to 0 (i.e. unlimited), but I tried 10000 and still
had complaints, so I bumped it down further to 2500, and it appears to be
running fine… Cheers…

O- ~ARK

David_Fletcher · December 13, 2001, 8:21pm

While looking at older messages about RT2 being slow I came accross this
thread. I feel really dumb reading the following example:

The other thing I’ve been doing to keep things running smoothly is running
the following statements nightly via cron – adapt as makes sense for you:

$dbh->do(“OPTIMIZE TABLE Tickets”);
print “Optimized Tickets.\n”;
$dbh->do(“OPTIMIZE TABLE Transactions”);
print “Optimized Transactions.\n”;
$dbh->do(“OPTIMIZE TABLE Attachments”);
print “Optimized Attachments.\n”;
$dbh->do(“OPTIMIZE TABLE Watchers”);
print “Optimized Watchers.\n”;
$dbh->do(“OPTIMIZE TABLE ObjectKeywords”);
print “Optimized ObjectKeywords.\nDone!\n\n”;

This clears out any deleted stuff and defrags the tables too I believe.

Could someone help out a dummy like me with an explanation of how to
implement this, or point me to the right documentation to figure it out?

Thanks,
David