Fedora Core 9 versus Ubuntu

Just an FYI email…

We have been using RT 3.6 for a year now. A few months ago we upgraded to 3.8 and ever since the performance on certain actions has been painful. In my opinion it is related to perl, not mysql. Sometimes I can submit a comment using the web interface and I will get the email before the screen refreshes.

So, on a whim I installed the latest version of Ubuntu and copied the database over. Oddly enough RT screams! So, I am going to start using Ubuntu instead of Fedora.

If anyone wants more details, I can do some version comparisons and such.

-Patrick

zbigniew@starpower.net wrote:

Just an FYI email…

We have been using RT 3.6 for a year now. A few months ago we upgraded to 3.8 and ever since the performance on certain actions has been painful. In my opinion it is related to perl, not mysql. Sometimes I can submit a comment using the web interface and I will get the email before the screen refreshes.

So, on a whim I installed the latest version of Ubuntu and copied the database over. Oddly enough RT screams! So, I am going to start using Ubuntu instead of Fedora.

If anyone wants more details, I can do some version comparisons and such.

Does anyone know if this is the problem:
http://blog.vipul.net/2008/08/24/redhat-perl-what-a-tragedy/
It would affect Centos5.2 as well.

Les Mikesell
lesmikesell@gmail.com

From: Les Mikesell lesmikesell@gmail.com
Subject: Re: [rt-users] Fedora Core 9 versus Ubuntu
To: zbigniew@starpower.net
Cc: rt-users@lists.bestpractical.com
Date: Wednesday, September 10, 2008, 2:36 PM
zbigniew@starpower.net wrote:

Just an FYI email…

We have been using RT 3.6 for a year now. A few
months ago we upgraded to 3.8 and ever since the performance
on certain actions has been painful. In my opinion it is
related to perl, not mysql. Sometimes I can submit a
comment using the web interface and I will get the email
before the screen refreshes.

So, on a whim I installed the latest version of Ubuntu
and copied the database over. Oddly enough RT screams! So,
I am going to start using Ubuntu instead of Fedora.

If anyone wants more details, I can do some version
comparisons and such.

Does anyone know if this is the problem:
http://blog.vipul.net/2008/08/24/redhat-perl-what-a-tragedy/
It would affect Centos5.2 as well.

Les,
In fc9, perl is OK again. There was an update to perl in fc8 that fixed the bless/overload bug. Only fc7 has that bug. As far as C5.2 goes, you can use the same SRPM that fixed fc8 and build it for Centos.

There has been alot of discussion about a CSS client bug in firefox that has slowed the rt3 web interface. I hate to say it, but log into rt3 using IE6+ and see if its faster!

Mark

For once things should be OK on the server side.

zbigniew@starpower.net wrote:

Just an FYI email…

We have been using RT 3.6 for a year now. A few months ago we upgraded to 3.8 and ever since the performance on certain actions has been painful. In my opinion it is related to perl, not mysql. Sometimes I can submit a comment using the web interface and I will get the email before the screen refreshes.

So, on a whim I installed the latest version of Ubuntu and copied the database over. Oddly enough RT screams! So, I am going to start using Ubuntu instead of Fedora.

If anyone wants more details, I can do some version comparisons and such.

Does anyone know if this is the problem:
http://blog.vipul.net/2008/08/24/redhat-perl-what-a-tragedy/
It would affect Centos5.2 as well.

Yes, this is huge problem. Several weeks ago I made a test on a clean
RT’s DB with stock RH’s perl and custom perl. RH’s was 10x slower and
with each request it gets slower and slower.


Les Mikesell
lesmikesell@gmail.com


http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-users

Community help: http://wiki.bestpractical.com
Commercial support: sales@bestpractical.com

Discover RT’s hidden secrets with RT Essentials from O’Reilly Media.
Buy a copy at http://rtbook.bestpractical.com

Best regards, Ruslan.

There has been alot of discussion about a CSS client bug in
firefox that has slowed the rt3 web interface. I hate to say it, but
log into rt3 using IE6+ and see if its faster!

Oh my god. RT is usable! Thank-you! I did an upgrade from 3.6.x to 3.8
and didn’t know why the UI experience was so terrible, but I was just
living with it and dreading opening long tickets.

Booting a WinXP VM and then use RT under IE, for me, is LESS LOAD than
using RT on FF on Linux (64-bit RedHat Desktop).

Thank-you again.

Brian

There has been alot of discussion about a CSS client bug in
firefox that has slowed the rt3 web interface. I hate to say it, but
log into rt3 using IE6+ and see if its faster!

Oh my god. RT is usable! Thank-you! I did an upgrade from 3.6.x to 3.8
and didn’t know why the UI experience was so terrible, but I was just
living with it and dreading opening long tickets.

Booting a WinXP VM and then use RT under IE, for me, is LESS LOAD than
using RT on FF on Linux (64-bit RedHat Desktop).

It’s all in the rounded corners. This is no good. “Looks sexy” isn’t
worth “doesn’t work.”

I don’t actually have any way to reproduce the issue myself. Do any of
you folks with the issue have any time to try to troll the firefox bug
tracker and/or interweb for anyone else who’s run into issues with
-moz-border-radius and Firefox on Linux?

There has been alot of discussion about a CSS client bug in
firefox that has slowed the rt3 web interface. I hate to say it, but
log into rt3 using IE6+ and see if its faster!

Oh my god. RT is usable! Thank-you! I did an upgrade from 3.6.x to 3.8
and didn’t know why the UI experience was so terrible, but I was just
living with it and dreading opening long tickets.

Booting a WinXP VM and then use RT under IE, for me, is LESS LOAD than
using RT on FF on Linux (64-bit RedHat Desktop).

It’s all in the rounded corners. This is no good. “Looks sexy” isn’t
worth “doesn’t work.”

I don’t actually have any way to reproduce the issue myself. Do any of
you folks with the issue have any time to try to troll the firefox bug
tracker and/or interweb for anyone else who’s run into issues with
-moz-border-radius and Firefox on Linux?

I found this via search for “firefox css performance”

https://bugs.launchpad.net/ubuntu/+source/firefox-3.0/+bug/223238

It sounds like FF is doing something with X, that the Nvidia drivers
don’t do well. That’s the only combination that seems to be consistent.
That’s the proprietary Nvidia drivers, not the opensource ones.

I don’t have Opera installed, but it sound like Opera doesn’t have the
same problem.

The hardware matches my setup, Dell Precision T7400 Dual Quad Xeon @
2.00 GHz/ 16 GB/ NVidia Quadro FX 1700 running RedHat 5.2 Desktop and
the proprietary NVidia drivers, and scrolling of a long RT Ticket really
hammers the machine.

Another result from "Firefox Nvidia scrolling’
http://www.nvnews.net/vbulletin/showthread.php?t=115790

Example page scrolls incredibly slow ( I have no idea what it says,
sorry if it’s offensive): http://www.tempus-vivit.net/taverne/thema/935

background:fixed ?

https://bugzilla.mozilla.org/show_bug.cgi?id=372039

position:fixed ? (open since 2003! Last update 2008-09-04)

https://bugzilla.mozilla.org/show_bug.cgi?id=201307

I couldn’t find anything specific to moz-border-radius, but now that
I’ve seeing so many example website perform so poorly under FF3, I think
I’ll be installing Opera.

Brian

Brian,

Thanks for the excellent triage. Does the solution mentioned at:

https://bugs.launchpad.net/ubuntu/+source/firefox-3.0/+bug/223238/comments/39

work for you?

-j