Hello,
I am doing some tests with SLES10 SP2 beta4 and an IBM DS4700 Storage
system, particularly focusing on DM-MP with the DS4000.
When configuring DM-MP I get inconsistent results when running multipath
with various options. I also have a problem with path fail back, which I
will describe in a separate mail to avoid confusion.
Here's the setup:
- IBM System x 3650 server with 2 QLA2460 HBAs
- SLES10 SP2 beta4 X86_64 installed on internal disk
- kernel-smp-2.6.16.60-0.6
- multipath-tools-0.4.7-34.32
- qla2xxx 8.02.00-k6-SLES10.2-02
- Brocade B32 4Gb switch, zoned that
each HBA can see one DS4700 controller
- DS4700 presenting a 50 GB LUN to the server
Another system is equipped with 2 Emulex LP1150 HBAs,
running lpfc 8.2.0.19. All the rest is the same as above.
I looked at the HW table of the multipath tools:
device {
vendor "IBM"
product "1814"
path_grouping_policy "group_by_prio"
path_checker "rdac"
hardware_handler "1 rdac"
prio "rdac"
no_path_retry "queue"
}
and decided that the internal defaults are ok for me with 2 exceptions:
- I want immediate automatic fail back
- I don't want "queue if no path".
So my "/etc/multipath.conf" is quite simple and looks like this:
x3650-lab-11:~ # cat /etc/multipath.conf
defaults {
no_path_retry 300
features "0"
failback immediate
user_friendly_names yes
}
blacklist {
devnode "^(ram|raw|loop|fd|md|dm-|sr|scd|st)[0-9]*"
devnode "^hd[a-z][[0-9]*]"
devnode "^cciss!c[0-9]d[0-9]*[p[0-9]*]"
wwid SServeRA_Daten_6673ADC0
wwid SServeRA_OS_77F79DC0
}
Here are the outputs of multipath with various options. Look at the
priority values and the [features=...] statement. BTW - the results are the
same on the machine with LPFC HBAs and also if I use the mpath_prio_rdac
callout.
Dry-Run
-------
x3650-lab-11:~ # multipath -d
create: mpatha (3600a0b800011fa78000033d947d7650d) n/a IBM,1814 FAStT
[size=50G][features=0][hwhandler=1 rdac]
\_ round-robin 0 [prio=2][undef]
\_ 4:0:0:0 sdd 8:48 [undef][ghost]
\_ round-robin 0 [prio=1][undef]
\_ 3:0:0:0 sdc 8:32 [undef][ready]
Configure DM-MP
---------------
x3650-lab-11:~ # multipath
create: mpatha (3600a0b800011fa78000033d947d7650d) n/a IBM,1814 FAStT
[size=50G][features=0][hwhandler=1 rdac]
\_ round-robin 0 [prio=2][undef]
\_ 4:0:0:0 sdd 8:48 [undef][ghost]
\_ round-robin 0 [prio=1][undef]
\_ 3:0:0:0 sdc 8:32 [undef][ready]
Show multipath topology (sysfs and DM info)
-------------------------------------------
x3650-lab-11:~ # multipath -l
mpatha (3600a0b800011fa78000033d947d7650d) dm-0 IBM,1814 FAStT
[size=50G][features=1 queue_if_no_path][hwhandler=1 rdac]
\_ round-robin 0 [prio=-1][active]
\_ 4:0:0:0 sdd 8:48 [active][undef]
\_ round-robin 0 [prio=-1][enabled]
\_ 3:0:0:0 sdc 8:32 [active][undef]
Show multipath topology (maximum info)
--------------------------------------
x3650-lab-11:~ # multipath -ll
mpatha (3600a0b800011fa78000033d947d7650d) dm-0 IBM,1814 FAStT
[size=50G][features=1 queue_if_no_path][hwhandler=1 rdac]
\_ round-robin 0 [prio=3][active]
\_ 4:0:0:0 sdd 8:48 [active][ready]
\_ round-robin 0 [prio=0][enabled]
\_ 3:0:0:0 sdc 8:32 [active][ghost]
The resulting device mapper table looks like this
x3650-lab-11:~ # dmsetup table mpatha
0 104857600 multipath 1 queue_if_no_path 1 rdac 2 1 round-robin 0 1 1 8:48
1000 round-robin 0 1 1 8:32 1000
With this said, here are my questions:
- How can I avoid that "queue if no path" is set?
- Am I using an outdated version of multipath?
- Which of the different outputs is the correct one?
- What is the true configuration (I can't "read" the dm table)?
- Are the differences in output a bug or are they correct and I can't
interpret them?
Many thanks in advance ........... Alex
Mit freundlichen Grüßen ... Kind regards ...
Alexander Warmuth
Advanced Technical Support System Storage Europe
Enterprise Disk High End Solutions
Mail: warmuth@xxxxxxxxxx
Phone: +49-6131-84-2887
Mobile: +49-171-5506083
Request ATS support at
http://web.mainz.de.ibm.com/e_dir/esccw3.nsf/ContentDocsByTitle/Request+support
IBM Deutschland GmbH, Vorsitzender des Aufsichtsrats: Hans Ulrich Maerki,
Geschäftsführung: Martin Jetter (Vorsitzender), Christian Diedrich,
Christoph Grandpierre, Matthias Hartmann, Thomas Fell, Michael Diemer, Sitz
der Gesellschaft: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB
14562, WEEE-Reg.-Nr. DE 99369940
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/dm-devel
