[WBEL-users] Grub, RAID and Tripwire
Mace.Scott@tatravelcenters.com
Mace.Scott@tatravelcenters.com
Wed, 11 Aug 2004 16:55:24 -0400
This is a multipart message in MIME format.
--=_alternative 0072ED3E85256EED_=
Content-Type: text/plain; charset="US-ASCII"
I have nearly 200 servers deployed with software RAID, and had to use lilo
instead of grub. Grub is very flaky with software RAID. I can't find the
forum articles that discuss this, it's been a while, but I ran into
trouble with drives failing, and the remaining drive (if sdb) is not
bootable. Lilo handled it perfectly. It was fairly easy to mass deploy
lilo, as all the boxes are identical. I'll see if I can track down the
discusssions regarding this.
=========================
Scott Mace
Systems Administrator
Travelcenters of America
440-808-4318
mace.scott@tatravelcenters.com
=========================
Alex Tkachenko <alex@ingrian.com>
Sent by: whitebox-users-admin@beau.org
08/11/2004 04:20 PM
To
whitebox-users@beau.org
cc
Subject
[WBEL-users] Grub, RAID and Tripwire
Hi Everybuddy,
Remember our discussion about grub not updating the slave MBR? I have
something to add to the story.
Yesterday I installed Tripwire, and run init, and check right away to
bring it to the settled state. But right on the first check it reported
that file /boot/grub/stage2 has different checksum. I commited the
report and after a couple of minutes run the check again. Guess what -
the checksum of the above file was different again. I repeated the
process several times and the checksum was flipping between two values.
My take is: the file IS different on two RAID members. Then I've checked
the md5sum of /usr/share/grub/i386-redhat/stage2 and /boot/grub/stage2.
Luckily, they were different :) Luckily because I might had it from the
second drive, where it is right and kept chasing it on and on.
OK, I have recalled that I updated manually the MBR on the second drive
a while ago. I went on and repeated the process for the FIRST drive.
This appears to bring the md5sum of the /boot/grub/stage2 in sync with
the one in /usr/share.
Has anybody seen this issue or had a different theory on the subject?
Please advise.
Thank you very much,
Alex
_______________________________________________
Whitebox-users mailing list
Whitebox-users@beau.org
http://beau.org/mailman/listinfo/whitebox-users
--=_alternative 0072ED3E85256EED_=
Content-Type: text/html; charset="US-ASCII"
<br><font size=2 face="sans-serif">I have nearly 200 servers deployed with
software RAID, and had to use lilo instead of grub. Grub is very
flaky with software RAID. I can't find the forum articles that discuss
this, it's been a while, but I ran into trouble with drives failing, and
the remaining drive (if sdb) is not bootable. Lilo handled it perfectly.
It was fairly easy to mass deploy lilo, as all the boxes are identical.
I'll see if I can track down the discusssions regarding this.</font>
<br><font size=2 face="sans-serif"><br>
=========================<br>
Scott Mace<br>
Systems Administrator<br>
Travelcenters of America<br>
440-808-4318<br>
mace.scott@tatravelcenters.com<br>
=========================</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>Alex Tkachenko <alex@ingrian.com></b>
</font>
<br><font size=1 face="sans-serif">Sent by: whitebox-users-admin@beau.org</font>
<p><font size=1 face="sans-serif">08/11/2004 04:20 PM</font>
<td width=59%>
<table width=100%>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td><font size=1 face="sans-serif">whitebox-users@beau.org</font>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td><font size=1 face="sans-serif">[WBEL-users] Grub, RAID and Tripwire</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><font size=2><tt>Hi Everybuddy,<br>
<br>
Remember our discussion about grub not updating the slave MBR? I have<br>
something to add to the story.<br>
<br>
Yesterday I installed Tripwire, and run init, and check right away to<br>
bring it to the settled state. But right on the first check it reported<br>
that file /boot/grub/stage2 has different checksum. I commited the<br>
report and after a couple of minutes run the check again. Guess what -<br>
the checksum of the above file was different again. I repeated the<br>
process several times and the checksum was flipping between two values.<br>
My take is: the file IS different on two RAID members. Then I've checked<br>
the md5sum of /usr/share/grub/i386-redhat/stage2 and /boot/grub/stage2.<br>
Luckily, they were different :) Luckily because I might had it from the<br>
second drive, where it is right and kept chasing it on and on.<br>
<br>
OK, I have recalled that I updated manually the MBR on the second drive<br>
a while ago. I went on and repeated the process for the FIRST drive.<br>
This appears to bring the md5sum of the /boot/grub/stage2 in sync with<br>
the one in /usr/share.<br>
<br>
Has anybody seen this issue or had a different theory on the subject?<br>
<br>
Please advise.<br>
<br>
Thank you very much,<br>
Alex<br>
<br>
<br>
<br>
<br>
_______________________________________________<br>
Whitebox-users mailing list<br>
Whitebox-users@beau.org<br>
http://beau.org/mailman/listinfo/whitebox-users<br>
</tt></font>
<br>
--=_alternative 0072ED3E85256EED_=--