I have seen this mentioned in another thread, but in playing around with the FreePBX 1.86.210.57-1 Beta Distro, no RAID is configured even when 2 hard drives are present.
Interestingly, I compared the kickstart file for the current release, 1.86.29.55, and the kickstart file for the 1.87.210.57-1 beta. It seems that the section of code that examines the hard drives and configures the software raid is exactly the same.
Yes it appears something has changed in Centos 5.7 and we have not had time to dig into the issue yet but we will get it resolved before we come out of Beta. If someone else has time to look into the issue and can figure it out that would be great also or it will have to wait tell after 2.10 and Astricon
I was looking in /var/log/anaconda.log and see and error running the command list-harddrives. Is it possible that the install is not able to determine how many drives are installed and does not create the file /tmp/part-include with the RAID config commands need to setup RAID?
I am doing another install and noticed that /tmp/part-include does not contain the commands to create the RAID array. It seems that the program /usr/bin/list-harddrives is failing. list-harddrives is a Python program and it is trying to import something called partedUtils and that import seems to be failing. If list-harddrives fails the the following code will not build the RAID array.
# Find the number of hard drives we have, if we find only one, no raid, more then one? raid.
harddrives=list-harddrives | wc -l
count=0
if [ “$harddrives” -gt “1” ]
then
for disk in list-harddrives | awk '{print $1}'
do
# Creates the BOOT partition that is 100mb in size
echo “part raid.0$count --size=100 --ondisk=$disk --asprimary” >> /tmp/part-include
# Creates the SWAP partition (1024mb)
echo "part raid.1$count --size=1024 --ondisk=$disk" >> /tmp/part-include
# Creates the root (/) partition that is able to grow to fill
echo "part raid.2$count --size=100 --grow --ondisk=$disk" >> /tmp/part-include
((count = count + 1))
done
((count = count - 1))
echo -n "raid /boot --level=1 --device=md0 --fstype=ext3 " >> /tmp/part-include
for disk in `seq 0 $count`
do
echo -n "raid.0$disk " >> /tmp/part-include
done
echo "" >> /tmp/part-include
echo -n "raid swap --level=1 --device=md1 --fstype=swap " >> /tmp/part-include
for disk in `seq 0 $count`
do
echo -n "raid.1$disk " >> /tmp/part-include
done
echo "" >> /tmp/part-include
echo -n "raid / --level=1 --device=md2 --fstype=ext3 " >> /tmp/part-include
for disk in `seq 0 $count`
do
echo -n "raid.2$disk " >> /tmp/part-include
done
echo "" >> /tmp/part-include
It looks like you are onto something here, alan_mousty. I am planning to do a little testing this afternoon to see if the suggestions in bugzilla will work. I will report my results.
We have a fix for this with the latest Beta Distro using Centos 5.7. If you install the 32 or 64 bit with more than 1 drive raid should be setup again.