Greetings - I’m looking into options for a ticketing system for our
helpdesk, and would very much appreciate any input.
Our call center gets around 3500 calls per week (increasing around 10% per
month). Today, we’re using a very thrown-together system that made sense
when we had 100 calls a day, but doesn’t any more. I’m looking into which
tool to implement, and hope to have something in place by July or August.
Can anyone share with me their reasons for selecting RT, as opposed to say
OTRS, or Remedy, or any of the dozens of commercial options? The managers
here are open to open-source (for the right reasons even, not just the
cost reason), which makes my life much easier when it comes to this sort
of a project.
The biggies we’re looking for aside from the normal email
handling/escalation/reporting stuff would be some way to tie in a
knowledge-base to the customer, so when they open a ticket, they select
what their problem is, and are offered a list of possible fixes. The goal
there is to help them right away, and to reduce call volume.
I’m very interested to hear your thoughts on how and why RT is the right
solution, if what I’ve described above makes sense for RT.
I’d also like suggestions on how to scale my hardware for a, let’s say,
10000 call/week volume. How “heavy” is this application? Does it scale
well to that volume?
Thanks,
Dave Hinz
Unix Systems Architect
eMagic.com, LLC
Quoting David Hinz David_Hinz@mgic.com:
Can anyone share with me their reasons for selecting RT, as opposed to say
OTRS, or Remedy, or any of the dozens of commercial options? The managers
here are open to open-source (for the right reasons even, not just the
cost reason), which makes my life much easier when it comes to this sort
of a project.
The biggies we’re looking for aside from the normal email
handling/escalation/reporting stuff would be some way to tie in a
knowledge-base to the customer, so when they open a ticket, they select
what their problem is, and are offered a list of possible fixes. The goal
there is to help them right away, and to reduce call volume.
Well, one good reason is that my company sunk $100k+ into Remedy just to
customize it enough to work, and it still doesn’t have the normal email
handling that RT comes with, free, no customization. And that was just for the
customizations; we still pay a site license fee, based on the # of users.
We also have support from BestPractical which is light-years better than the
“support” we get from the Remedy folks.
I can’t guarantee that RT is scaleable, but my impression is that it’s largely
dependent on your architecture. Most of the bottleneck is CPU and I/O with the
web server and the database, at least from what I’ve seen on the list.
“SK” == Sheeri Kritzer sheeri.kritzer@tufts.edu writes:
SK> I can’t guarantee that RT is scaleable, but my impression is that
SK> it’s largely dependent on your architecture. Most of the
SK> bottleneck is CPU and I/O with the web server and the database, at
SK> least from what I’ve seen on the list.
I would find it difficult to think you could peg a web server CPU
under even heavy load unless you’re running RT as a plain-jane CGI.
With a properly architected mod_perl or fast/speedy cgi setup, a
single pentium 3 class machine with 512Mb RAM should be able to handle
hundreds of requests per minute.
The bottleneck in this application is the database. Having well-tuned
database (software and hardware) is vital.
If your database is going to get HUGE, go with PostgreSQL as it seems
to scale better in my experience. Also, avoid linux with its 2Gb file
size limit, which will really cause you pain (especially if you’re
using mysql).
Naturally, someone will say, “but that’s fixed in version xyz of
mysql”. I don’t care. 
Vivek Khera, Ph.D. Khera Communications, Inc.
Internet: khera@kciLink.com Rockville, MD +1-240-453-8497
AIM: vivekkhera Y!: vivek_khera http://www.khera.org/~vivek/