Moving RT from sparc to x86 OS on VMWare?

We are looking at moving our RT server from a SPARC Sunblade 150 running Solaris 9 over to an x86 OS on a VMWare ESX virtual infrastructure node.

Anyone done this?

Any comments?

I’m looking to answer a few things:

Which OS would I be better off using in terms of ease of setup and performance / cost / support:

  • Fedora Core 3
  • Red Hat Enterprise Linux ES release 4 (update 1)
  • Solaris 10 x86

Any ideas on how to migrate from on RT release (3.0.3) on SPARC over to the new OS / platform and RT 3.4.2?

I was thinking:

  • Install a minimal OS on the new Virtual node
  • Satisfy RT 3.4.2 dependencies
  • Install and configure 3.0.3 on the new virtual node
  • Copy the existing 3.0.3 database over and test
  • Upgrade to 3.4.2
  • Cut over the from the SPARC to the new virtual node

Any comments / suggestions?
The information contained in this e-mail message is PRIVATE. It may contain confidential information and may be legally privileged. It is intended for the exclusive use of the addressee(s). If you are not the intended recipient, you are hereby notified that any dissemination, distribution or reproduction of this communication is strictly prohibited. If the intended recipient(s) cannot be reached or if a transmission problem has occurred, please notify the sender immediately by return e-mail and destroy all copies of this message.
Thank you.

Randy Millis (lists account) wrote:

We are looking at moving our RT server from a SPARC Sunblade 150 running
Solaris 9 over to an x86 OS on a VMWare ESX virtual infrastructure node.

Anyone done this?

Any comments?

I’m looking to answer a few things:

Which OS would I be better off using in terms of ease of setup and
performance / cost / support:

  • Fedora Core 3
  • Red Hat Enterprise Linux ES release 4 (update 1)
  • Solaris 10 x86
    Which one you familiar with.
    I don’t like RH based distros(IMHO), but there is good InstallGuide for
    RHEL4 on the wiki.

Any ideas on how to migrate from on RT release (3.0.3) on SPARC over to
the new OS / platform and RT 3.4.2?

I was thinking:

  • Install a minimal OS on the new Virtual node
  • Satisfy RT 3.4.2 dependencies
  • Install and configure 3.0.3 on the new virtual node
  • Copy the existing 3.0.3 database over and test
  • Upgrade to 3.4.2
  • Cut over the from the SPARC to the new virtual node
    Roadmap looks good.

Which OS would I be better off using in terms of ease of setup and
performance / cost / support:

  • Fedora Core 3
  • Red Hat Enterprise Linux ES release 4 (update 1)
  • Solaris 10 x86
    Which one you familiar with.
    I don’t like RH based distros(IMHO), but there is good InstallGuide for
    RHEL4 on the wiki.

If you don’t need vendor support, Centos4 http://www.centos.org/ is a
free alternative that is rebuilt from the RHEL4 source RPMS. Fedora
will work but the fast release cycle makes it hard to keep a server up
to date.

Les Mikesell
lesmikesell@gmail.com

Hi,
Solaris x86 is great for RT and as you are already using Solaris on
sparc this would be the best solution.
Solaris 10 is also FREE :slight_smile:
Just pick a machine that meets the HCL

Les Mikesell wrote:>On Thu, 2005-07-14 at 04:26, Ruslan U. Zakirov wrote:

Which OS would I be better off using in terms of ease of setup and
performance / cost / support:

  • Fedora Core 3
  • Red Hat Enterprise Linux ES release 4 (update 1)
  • Solaris 10 x86

Which one you familiar with.
I don’t like RH based distros(IMHO), but there is good InstallGuide for
RHEL4 on the wiki.

If you don’t need vendor support, Centos4 http://www.centos.org/ is a
free alternative that is rebuilt from the RHEL4 source RPMS. Fedora
will work but the fast release cycle makes it hard to keep a server up
to date.

Cheers

Richard Skelton
Richard.Skelton@infineon.com
Infineon Technologies UK Ltd
Infineon House
Great Western Court
Hunts Ground Road
Stoke Gifford
Bristol
BS34 8HP
Tel +44(0)117 9528808

Which OS would I be better off using in terms of ease of setup and
performance / cost / support:

  • Fedora Core 3
  • Red Hat Enterprise Linux ES release 4 (update 1)
  • Solaris 10 x86

Fedora core is not the answer for any server unless you like the idea of
doing a distribution upgrade to some new fedora core in less than a year.

RHEL (or Centos) is a better choice.

Debian is not a horrible choice either if you have experience with it.
RT has lots of dependencies that can take forever to build by hand.
debian’s packaging system makes it easy. i say this not being a big
debian fan in general.

as far as solaris vs. linux, depends mostly on your own environment.
but you’ll probably have to go build more stuff “by hand” with solaris
vs. redhat. (tough call - check sunfreeware.com, and the “DAG” rpm
repository for redhat if you want binary distributions)

danno
dan pritts - systems administrator - internet2
734/352-4953 office 734/834-7224 mobile

Richard Skelton wrote:

Hi,
Solaris x86 is great for RT and as you are already using Solaris on
sparc this would be the best solution.
Solaris 10 is also FREE :slight_smile:
Just pick a machine that meets the HCL
Hardware | Oracle

Big deal - he uses VMware ESX-server.
That pretty much has his hardware defined :wink:
The only variables are the amount of virtual RAM and diskspace it gets,
together with the number of virtual NICs.

As a VMware ESX-customer (usually not under 25k USD, but most probably
in the 6-figure range, depending on the scale), I’d got straight to
VMware Inc. and demand that FreeBSD5 starts working on it ASAP.
(It works on GSX, so it shouldn’t be that hard to get working)

Both from a RT and from a Solaris-admin perspective, this will probably
make the most sense.

FreeBSD4.9 is certified, but I would only very hesitantly install a
server with, now that 6.0 has been branched and will be released by the
end of the year…

GSX has experimental support for both Solaris 10 and full support for
FreeBSD5 - so you could start with a GSX-vm and later, when VMware Inc.
gets their act together, move it to the GSX-server.
But as someone else said: with Solaris, you’ll have to roll-your-own for
pretty much any package and perl-module - and I counted (I think) 86
p5-packages on my FreeBSD5-box after RT was installed…
On RHEL, there’s no guarantee that a future update will not break the
whole thing - unless you also do a roll-your-own install of everything
that is needed by RT…

cheers,
Rainer

Thanks to everyone for the comments on my question.

I am in the process of trying an eval of Red Hat Enterprise Linux ES release 4 (update 1) and then Solaris 10 x86 to see if there are any performance differences or ease of installation / maintenance issues on either OS.----- Original Message -----
From: Randy Millis (lists account)
To: rt-users@lists.bestpractical.com
Sent: Wednesday, July 13, 2005 9:56 AM
Subject: [rt-users] Moving RT from sparc to x86 OS on VMWare?

We are looking at moving our RT server from a SPARC Sunblade 150 running Solaris 9 over to an x86 OS on a VMWare ESX virtual infrastructure node.

Anyone done this?

Any comments?

I’m looking to answer a few things:

Which OS would I be better off using in terms of ease of setup and performance / cost / support:

  • Fedora Core 3
  • Red Hat Enterprise Linux ES release 4 (update 1)
  • Solaris 10 x86

Any ideas on how to migrate from on RT release (3.0.3) on SPARC over to the new OS / platform and RT 3.4.2?

I was thinking:

  • Install a minimal OS on the new Virtual node
  • Satisfy RT 3.4.2 dependencies
  • Install and configure 3.0.3 on the new virtual node
  • Copy the existing 3.0.3 database over and test
  • Upgrade to 3.4.2
  • Cut over the from the SPARC to the new virtual node

Any comments / suggestions?
The information contained in this e-mail message is PRIVATE. It may contain confidential information and may be legally privileged. It is intended for the exclusive use of the addressee(s). If you are not the intended recipient, you are hereby notified that any dissemination, distribution or reproduction of this communication is strictly prohibited. If the intended recipient(s) cannot be reached or if a transmission problem has occurred, please notify the sender immediately by return e-mail and destroy all copies of this message.
Thank you.

The rt-users Archives

Be sure to check out the RT Wiki at http://wiki.bestpractical.com

Solaris x86 is not supported under ESX (as of 2.5.1 at least)-- in fact
after much work trying to get it somehow to work anyway, I gave up.

If you are looking at ESX, I’d pick CentOS 4.1 (RHEL4U1 equivalent).
Make sure you use the LSI rather than BusLogic SCSI driver (there isn’t
a BusLogic driver by default in the RHEL4/Centos4 release).

I am running Debian with Exim4, Postgres and Apache2. Everything works fine
but when I send an email to the RT system it won’t process and create a
ticket. I can see in the Exim4 log that it’s trying to run the pipe
command:

|/usr/bin/rt-mailgate --queue ‘Incident Reports’ --action correspond --url
http://localhost

But receive this error:

pipe to |/usr/bin/rt-mailgate --queue ‘Incident Reports’ --action correspond
–url http://localhost
generated by rtir@xx.txstate.edu
Delay reason: pipe_transport unset in system_aliases router

I have done extensive searching on this list and elsewhere and can’t find a
solution.

Thanks in advance for any help!

Elliott Franklin, CISSP
Information Security Officer
Texas State University - San Marcos
512.245.2501

Have you seen this article:
http://mail.python.org/pipermail/mailman-users/2005-February/042410.html

http://wiki.bestpractical.com/index.cgi?search_term=pipe_transport&action=search

http://www.google.com/search?q=Delay+reason%3A+pipe_transport+unset+in+system_aliases+routerOn Thu, 14 Jul 2005, IT Security wrote:

I am running Debian with Exim4, Postgres and Apache2. Everything works fine
but when I send an email to the RT system it won’t process and create a
ticket. I can see in the Exim4 log that it’s trying to run the pipe
command:

|/usr/bin/rt-mailgate --queue ‘Incident Reports’ --action correspond --url
http://localhost

But receive this error:

pipe to |/usr/bin/rt-mailgate --queue ‘Incident Reports’ --action correspond
–url http://localhost
generated by rtir@xx.txstate.edu
Delay reason: pipe_transport unset in system_aliases router