Raid Problems rh6.1 - 2.2.14 kernel HP-LXr8500 with a NetRaid Card
Haniel G. Mercedes
I just installed Linux rh6.1 running 2.2.14 kernel,
so I am trying to configure a Raid 0 using raid tools v0.90,
but I am having some problems;
According to
http://sunsite.utk.edu/ftp/pub/linux/slackware/slackware/docs/Linux-HOWTO/HP
-HOWTO
there is an info that says:
For the RedHat 6.1 distribution, the MegaRaid driver provided
By the kernel does support only the first logical drive. In order
to obtain the others, please recompile a 2.2.14 standard kernel.
If you want to obtain that info, search at the document for "-DHP", the
next paragraph will be the info.
I have 2 types of problems.
Problem 1:
I upgraded from 2.2.12 to 2.2.14 Kernel but I could not find a raid patch
for 2.2.14 Kernel,
the last recent patch that I got, was raid0145-19990824-2.2.11.tar.gz, so
I did not patch the kernel with this old patch kernel.
Anyway with 2.2.14 kernel, I can see all raid disks from my raid
controller, so I sliced all disk as a "Linux raid" type "fd",
also made the raids partitions, but when I tried to use the raid tools, I
could not be able to define a raid 0,
I really think that I defined the /etc/raidtab file correctly and my
/proc/mdstat looks like is good.
Report Problem:
[root@puma /root]# mkraid /dev/md0
handling MD device /dev/md0
analyzing super-block
disk 0: /dev/sdc1, 17775891kB, raid superblock at 17775808kB
disk 1: /dev/sdd1, 17775891kB, raid superblock at 17775808kB
disk 2: /dev/sde1, 17775891kB, raid superblock at 17775808kB
disk 3: /dev/sdf1, 17775891kB, raid superblock at 17775808kB
disk 4: /dev/sdg1, 17775891kB, raid superblock at 17775808kB
mkraid: aborted, see the syslog and /proc/mdstat for potential clues.=====>
(This is the error)
[root@puma /root]# cat /proc/mdstat
Personalities: [1 linear] [2 raid0]
read_ahead not set
md0: inactive
md1: inactive
md2: inactive
md3: inactive
[root@puma /root]# cat /etc/raidtab
raiddev /dev/md0
raid-level 0
persistent-superblock 1
chunk-size 8
nr-raid-disks 5
nr-spare-disks 0
device /dev/sdc1
raid-disk 0
device /dev/sdd1
raid-disk 1
device /dev/sde1
raid-disk 2
device /dev/sdf1
raid-disk 3
device /dev/sdg1
raid-disk 4
I just looked at /var/log/dmesg and dmesg command and there is not problem
with this.
I just upgraded the Server BIOS to last version from any HW problem.
(NetRaid Card)
Poblem 2:
First:
I tried with a new kernels versions like 2.2.15 or 2.2.16, 2.2.17, but
when I try to boot with all of these new kernels the system hang. I tried
with the rh6.2 and rh 7.0 CD and I got the same error.
This is the mesg that appear at the screen monitor:
===================================================================
blah ......
blah blah......
blah blah blah.......
PCI: PCI BIOS revision 2.16
entry at 0xfdaca
PCI: Using configuration type 1
PCI: probing pci hardware
PCI: 00:c0[0e11/6010] Scanning per hosts bridge
PCI: Scanning compaq per hosts bridge 18/00
PCI: 00:c8[0e11/6010] Scanning compaq per hosts bridge
PCI: Scanning compaq per hosts bridge 19/00
PCI: 00:d0[0e11/6010] Scanning compaq per hosts bridge
PCI: Scanning compaq per hosts bridge 1a/00
After this the system HANGS.
===================================================================
Basically in this case, I think that the problem is when rh6.2 is probing
the PCI buses, its hang probing the third PCI bus SCSI at the NetRaid
controller.
I could boot with rh6.2 CD or rh7.0 if I remove the NetRaid card from the
server HP-LXr8500. I am not expecting that someone is going to tell me the
solution of this problem, cause perhaps it could be "hardware problem", the
reason is that I am writing this, is because I hope that someone got a
similar situacion and maybe could get me some ideas, I am very interests in
to use that NetRaid Card, is really pretty fast.
===================================================================
Questions:
- Any idea please?
- Is there a new raid patch for 2.2.(14,15,16,17) kernel ?
- I suppose that I have the raid defined correctly, what does mean "mkraid:
aborted, see the syslog and /proc/mdstat for potential clues" ?
Thanks,
Ps. Sorry for this large doc and sorry for an especific hardware question.
Haniel Mercedes