Connecting Red Hat 5.3 to EVA 8100 vdisk

[J Borgia]

Okay folks, bear with me...I'm a Solaris admin playing in Red Hat land today. The instructions for connecting a Solaris system to the EVA via MPIO where very straight forward. I'm not finding the same experience with Red Hat 5.3 Can anyone point me to a useful resource for seeing the Vdisk in Red Hat Linux, then setting up the multipathing? I've found one document that says I should use HPDM for Linux, but it doesn't list the EVA 8100 as a supported platform. Then I saw another that says I should use the bundled RH Linux multiplathing software, which seems less than straight forward.

Can anyone give me a hand with this? Thanks.

Joe

[Hakki Aydin Ucar]

There is not too much topic:
Here I can find:

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1025366

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=960206

[Fayez · Accepted solution]

Hi Joe,

Please find this link, hope it will help you.

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=18964&prodSeriesId=3559651&swItem=co-71367-1&prodNameId=3559652&swEnvOID=4035&swLang=13&taskId=135&mode=3

Zalloum

[Rob Leadbeater]

Hi Joe,

This doc should also be useful:

http://bizsupport2.austin.hp.com/bc/docs/support/SupportManual/c01080211/c01080211.pdf

Cheers,

Rob

[J Borgia]

I've got what looks like a device that can be mounted as /dev/mapper/multipath1, but for some weird reason, it fails when I put it in the fstab. Any hints?

[Rob Leadbeater]

Seeing what you're actually doing, and the actual error messages might be useful...

Cheers,

Rob

[westb]

maybe a stupid question...you made a filesystem on it right ?

For what it's worth, we have found the RH native multipathing to be very solid and we are using an EVA 8100 with RHEL5. There are a lot of underlying devices but if you use friendly names configured in /etc/mutlipath.conf it's easier to deal with and you don't have to worry about devices that change across reboots.

The steps that I use:
1. create vdisk on EVA
2. present it to a server
3. Copy and paste the WWN of the new vdisk into /etc/multipath.conf on the server (see sample below).
4. On the server, scan for the new disks (this is a 2 port Qlogic card):
echo "- - -" > /sys/class/scsi_host/host0/scan
echo "- - -" > /sys/class/scsi_host/host1/scan
5. run multipath -ll to see the new disks and multipath device
6. mkfs on the multipath device
7. mount the filesystem

hope this helps.

-----------------------
sample multipath.conf:
multipaths {
multipath {
wwid 3600508b40006af2c0000f000033e0000
alias data2
}
}

devnode_blacklist {
devnode "^(ram|raw|loop|fd|md|dm-|sr|scd|st)[0-9]*"
devnode "^hd[a-z]"
devnode "^cciss!c[0-9]d[0-9]*"
}

defaults {
udev_dir /dev
polling_interval 10
selector "round-robin 0"
path_grouping_policy failover
getuid_callout "/sbin/scsi_id -g -u -s /block/%n"
prio_callout "/bin/true"
path_checker tur
rr_min_io 100
rr_weight uniform
failback immediate
no_path_retry 12
user_friendly_names yes
}

devices {

device {
vendor "HP|COMPAQ"
product "HSV1[01]1 \(C\)COMPAQ|HSV2[01]0"
path_grouping_policy group_by_prio
getuid_callout "/sbin/scsi_id -g -u -s /block/%n"
path_checker tur
path_selector "round-robin 0"
prio_callout "/sbin/mpath_prio_alua /dev/%n"
rr_weight uniform
failback immediate
hardware_handler "0"
no_path_retry 12
}

device {
vendor "HP"
product "OPEN-.*"
path_grouping_policy multibus
getuid_callout "/sbin/scsi_id -g -u -s /block/%n"
path_selector "round-robin 0"
rr_weight uniform
prio_callout "/bin/true"
path_checker tur
hardware_handler "0"
failback immediate
no_path_retry 12
}

}

[J Borgia]

Westb,

So I have followed your instructions here. When I run multipath -ll, this is the output:

mpath1 (36001438002a55d780000800000590000) dm-6 HP,HSV210
[size=500G][features=1 queue_if_no_path][hwhandler=0][rw]
\_ round-robin 0 [prio=200][active]
\_ 3:0:0:1 sdb 8:16 [active][ready]
\_ 3:0:2:1 sdd 8:48 [active][ready]
\_ 4:0:0:1 sdf 8:80 [active][ready]
\_ 4:0:1:1 sdg 8:96 [active][ready]
\_ round-robin 0 [prio=30][enabled]
\_ 3:0:1:1 sdc 8:32 [active][ready]
\_ 3:0:3:1 sde 8:64 [active][ready]
\_ 4:0:2:1 sdh 8:112 [active][ready]

Do I run mkfs on "mpath" as the device? What I was doing before was:

mkfs.ext3 /dev/mapper/mpath1

It would build a filesystem with no complaints. I could even then mount it at /mnt as a test.

When I put it in the /etc/fstab for mounting at boot time, it will hang in the boot for not being able to recognize the device. It's almost like somehow the multipath functions have a service that seems to start AFTER mounting at boot, but I just don't know.

So that's where I am, I know if I mkfs the mpath1 device, it'll hang me at boot and I'll need to go into rescue mode to remove the line from fstab.

Any ideas?

[westb]

Is it just a mounting issue ? i.e. after removing the line in fstab and rebooting you can then manually mount it ?

If the suggestion below isn't the problem then post the message you receive at boot time.

In /etc/fstab it should look like:
/dev/mapper/mpath1 /data ext3 defaults 1 2

[westb]

Post your fstab file too.

[J Borgia]

I tried to enter the mpath1 device into the fstab as you described, but it also failed. I had to fix the fstab again after reboot. When I did, I logged back in and looked into /dev/mapper and noticed that mpath1 was now replaced with the alias I set in the multipath.conf file: frydata1. I was able to manually mount /dev/mapper/frydata1 to /mnt, so I decided to give this a try in fstab. This seemed to work.

After two reboots, it automatically now mounts my SAN disk with that device alias in the fstab.

Thanks for the help all.

[J Borgia]

Thanks everyone for your help with this.