|
List Info
Thread: Kickstart issue
|
|
| Kickstart issue |
 
United States |
2007-06-01 12:27:50 |
|
I have an issue re-kickstarting RHEL 4 servers when I need to redploy
them. I have HP Linux servers that the CCISS raid controller (came out of
Compaq). The disk presented by this raid controller should up as /dev/cciss/c#d#.
If I have the systems connected to SAN storage then they show up as /dev/sd[a,
b,c, etc.]. If I try to kickstart a HP server that has SAN devices already
zoned to it the kickstart process fails with “Input/output error during
write on /dev/sda” error. Does anyone know what could be the
problem and how to address. I’m would like to eleminate having to
disconnecting the fiber connections or unzoning luns.
Here’;s my partition section of the ks.cfg
zerombr yes clearpart --drives=cciss/c0d0,cciss/c0d1 --all part swap ; ; --start=1 --end=800 --ondisk cciss/c0d0 --asprimary part / --fstype ext3 --start=801 --end=1057 --ondisk cciss/c0d0 --asprimary part pv.00 --size 1 --grow --asprimary --ondisk cciss/c0d0 volgroup vg00 pv.00 logvol /usr --vgname=vg00 --size=2024 --name=lvol1 logvol /var --vgname=vg00 --size=4096 --name=lvol2 logvol /home --vgname=vg00 --size=1024 --name=lvol3 logvol /opt --vgname=vg00 --size=4096 --name=lvol4
Thanks for any help,
|
| RE: Kickstart issue |
United States |
2007-06-01 12:44:32 |
|
|
This is a problem for us as well. ; The most reliable
solution I've been given (but haven't had time to try) is to bust open your
initrd and change the entry for your HBA to "ignore" in the /modules/pcitable
file.
There are a number of options that are SUPPOSED to work,
like latefcload, but it hasn't worked for us.
From:
kickstart-list-bounces redhat.com [mailto:kickstart-list-bouncesredhat.com]
On Behalf Of Robbie Pike
Sent: Friday, June 01, 2007 10:28
AM
To: kickstart-listredhat.com
Subject: Kickstart
issue
I have an issue re-kickstarting RHEL 4 servers when I
need to redploy them. I have HP Linux servers that the CCISS raid controller
(came out of Compaq). The disk presented by this raid controller should up as
/dev/cciss/c#d#. If I have the systems connected to SAN storage then they show
up as /dev/sd[a, b,c, etc.]. If I try to kickstart a HP server that has SAN
devices already zoned to it the kickstart process fails with “Input/output error
during write on /dev/sda” error. Does anyone know what could be the
problem and how to address. I’m would like to eleminate having to disconnecting
the fiber connections or unzoning luns.
Here’;s my partition section of the
ks.cfg zerombr yes clearpart --drives=cciss/c0d0,cciss/c0d1 --all part swap ; ; --start=1 --end=800 --ondisk cciss/c0d0 --asprimary part / --fstype ext3 --start=801 --end=1057 --ondisk cciss/c0d0 --asprimary part pv.00 --size 1 --grow --asprimary --ondisk cciss/c0d0 volgroup vg00 pv.00 logvol /usr --vgname=vg00 --size=2024 --name=lvol1 logvol /var --vgname=vg00 --size=4096 --name=lvol2 logvol /home --vgname=vg00 --size=1024 --name=lvol3 logvol /opt --vgname=vg00 --size=4096 --name=lvol4
Thanks for any
help, |
| Re: Kickstart issue |
|
2007-06-04 19:55:49 |
On 6/1/07, Shabazian, Chip <Chip.Shabazianbankofamerica.com> wrote:
>
>
> This is a problem for us as well. The most reliable
solution I've been
> given (but haven't had time to try) is to bust open
your initrd and change
> the entry for your HBA to "ignore" in the
/modules/pcitable file.
>
> There are a number of options that are SUPPOSED to
work, like latefcload,
> but it hasn't worked for us.
latefcload was added in a fairly recent update for RHEL 3,
but I don't
know if it's in RHEL 4. "nostorage" is another
option only for RHEL 4
(and probably RHEL 5, but I have not tested it). On my
previous team,
we specified "nostorage" for all servers and then
ks.cfg contained
"device scsi cciss" for HP. This made sure we
didn't overwrite any
SAN storage.
/Brian/
_______________________________________________
Kickstart-list mailing list
Kickstart-listredhat.com
https://www.redhat.com/mailman/listinfo/kickstart-list
|
|
| Re: Kickstart issue |
United States |
2007-12-18 09:06:33 |
We ran into the same issue at my company, and I didn't want
anaconda to install
on the SAN and wipe out a database. The trick is to use a
python script in the
%pre section of your kickstart to write out a partitioning
file, then import
that using %import. Your kickstart should looks like so:
install
text
network --bootproto dhcp
...
zerombr yes
bootloader --location mbr
clearpart --all --initlabel
%include /tmp/part-include
%packages
...
%pre
/usr/bin/python <<EOB
import sys
sys.path.append("/usr/lib/anaconda")
import isys
drive = isys.hardDriveDict().keys()[0]
f = open("/tmp/part-include", "w")
f.write("partition /boot --fstype=ext3 --size=200
--ondisk=%sn" % (drive,))
f.write("partition / --fstype=ext3 --size=10240
--ondisk=%sn" % (drive,))
f.write("partition /var --fstype=ext3 --size=4096
--ondisk=%sn" % (drive,))
f.write("partition swap --size=2048 --ondisk=%sn"
% (drive,))
f.write("partition swap --size=2048 --ondisk=%sn"
% (drive,))
f.write("partition pv.01 --size=1 --grow
--ondisk=%sn" % (drive,))
f.write("volgroup vg0 pv.01 --pesize=32768n")
f.write("logvol /logs --vgname vg0 --size=1 --grow
--name logs --fstype=ext3n")
f.close()
EOB
%post
..
_______________________________________________
Kickstart-list mailing list
Kickstart-listredhat.com
https://www.redhat.com/mailman/listinfo/kickstart-list
|
|
[1-4]
|
|
|
about | contact Other archives ( Real Estate discussion Medical topics )
|