Upgrade from 11.1 to 11.2 on RHEL5

[payless]

I am going to upgrade from 11.1 to 11.2. What is the best process to follow? Are there any things I should be careful of?

Regards
Kerry

[jaime.pinto]

This is a redundant post. Please READ or SEARCH the forum before opening new threads and contribute to the existing ones.
viewtopic.php?t=8893

[payless]

This is a redundant post. Please READ or SEARCH the forum before opening new threads and contribute to the existing ones.
viewtopic.php?t=8893

Hi Jaime,
I have READ and SEARCHED the forum before I opened a new thread.
I have looked at the post you refer to a number of times and it does not answer my question.
I still need to know the best process to follow and if there are any issues I need to be aware of. The post that you refer to does not address my first question at all.

[jaime.pinto]

I still think you could have asked/added your question within the context of the original thread I referred you to. This way the next guy that comes in with questions on 11.1 to 11.2 upgrade can find information under one single location, even if they relate to multiple OS's.

We did a 11.1 to 11.2 on RHEL5 upgrade this past week without any *new* problems. Remember to do a omshut and stop the DB and tomcat before running the install script. Afterwards reboot the server, just to be sure all services come up as they should.

[payless]

Hi Jaime,

Thanks for the feedback.

[konman]

Jaime,

I'm in the same position as payless - I have the jumped into the community edition and I was hoping that there may have been a small wiki entry of step by step instructions such as the specific code of stopping the DB and stopping tomcat.

Once I find what has to be run I will post it back here for the next next guy.

Cheers,
KonMan.

[gren]

I have an "sx.stop" script I use for stopping Scalix, which you may want to review and adapt for your own purposes.

Code: Select all

#!/bin/bash

sc_name=`basename $0`

if  [ $(id -u) -ne 0 ]
then
        echo "$sc_name: Must be run as root - FAILED"
        exit 1
fi

for initscript in scalix-tomcat scalix-postgres sendmail scalix
do
    if [ -f /etc/init.d/$initscript ]
    then
        (set -x; /etc/init.d/$initscript stop)
    fi
done

# look for tricky process and kill them if present.
for baddy in omnsdm omnssck omdbmon omlicmon omctmon omissdm cda.server \
             omsmdm omdrs ual.remote ual.local license.server ldapmapper \
             omsmtpd in.imap41d pop3.server omslapdeng index.catcher \
             advmail.sckd omslapd postmaster
do
    /sbin/pidof $baddy >/dev/null
    if [ $? -eq 0 ]
    then
        (set -x;/sbin/pidof $baddy)
        pids=$(/sbin/pidof $baddy)
        (set -x;kill ${pids})
        (set -x;sleep 4)
        pids=$(/sbin/pidof $baddy)
        [ $? -eq 0 ] && (set -x;kill -9 ${pids})
    fi
done

numSxUsers=$(ps -u scalix --format %p|grep -v PID|wc -l)
if [ $numSxUsers -gt 0 ]
then
        (set -x;ps -u scalix)
        (kill `ps -u scalix --format %p|grep -v PID`)
        (set -x;sleep 4)
        (kill -9 `ps -u scalix --format %p|grep -v PID`)
fi

ipcs -a|grep scalix >/dev/null 2>&1
if [ $? -eq 0 ]
then
    echo 'IPCS system still in use for Scalix - wait for things to settle'
    (set -x;sleep 20)
fi

for IPCS_CLASS in q m s
do
    (set -x;ipcs -${IPCS_CLASS}|grep scalix)
    if [ $? -eq 0 ]
    then
        echo "Cleaning up Scalix IPCS usage for class $IPCS_CLASS"
        # message queues don't necessarily disappear straight away.
        (set -x;sleep 15)
        echo '>>> ps -flp 999 `sxprocs`'
        ps -flp 999 `sxprocs`
        if [ "$(sxprocs)" != "" ]
        then
            (set -x;kill ${pids})
            (set -x;sleep 5)
        fi
        (set -x;sleep 2)
        ipcs -${IPCS_CLASS}|grep scalix|awk '{printf "(set -x;ipcrm -'"${IPCS_CLASS}"' %s)\n", $2}'|bash
    fi
done

exit 0


You may not strictly need to stop sendmail but it may potentially do LDAP lookups which might fail if done at the wrong moment.

Hope this helps,
Gren