Greetings,

I perform SAN-based snapshots on OCFS2 and ext3 filesystems under
dm-multipath. Before I snapshot the filesystem, I unmount and delete
the previous night's snapshot LUNs. Doing so causes (correctly)
multipathd to output the following errors:

<snip>
...
Jul 14 03:16:02 ausracdbd01 multipathd: mpath19: remaining active paths:
7
Jul 14 03:16:02 ausracdbd01 multipathd: sdci: tur checker reports path
is down
Jul 14 03:16:02 ausracdbd01 kernel: device-mapper: multipath: Failing
path 69:80.
Jul 14 03:16:02 ausracdbd01 multipathd: sdco: tur checker reports path
is down
....
</snip>

During this snapshot time, I would like to temporarily turn off the
verbosity of multipathd. Is this possible? It's mostly an annoyance
factor for me because our nightly centralized syslog scan picks up words
like 'fail' and generates a report to the system administrators (which
we then have to ignore for these messages).

TIA,

Daniel


__________________________________________________ ____________________
This email transmission and any documents, files or previous email
messages attached to it may contain information that is confidential or
legally privileged. If you are not the intended recipient or a person
responsible for delivering this transmission to the intended recipient,
you are hereby notified that you must not read this transmission and
that any disclosure, copying, printing, distribution or use of this
transmission is strictly prohibited. If you have received this transmission
in error, please immediately notify the sender by telephone or return email
and delete the original transmission and its attachments without reading
or saving in any manner.


--
dm-devel mailing list
dm-devel@redhat.com
https://www.redhat.com/mailman/listinfo/dm-devel

> Greetings,
>
> I perform SAN-based snapshots on OCFS2 and ext3 filesystems under
> dm-multipath. Before I snapshot the filesystem, I unmount and delete
> the previous night's snapshot LUNs. Doing so causes (correctly)
> multipathd to output the following errors:
>
> <snip>
> ...
> Jul 14 03:16:02 ausracdbd01 multipathd: mpath19: remaining active paths:
> 7
> Jul 14 03:16:02 ausracdbd01 multipathd: sdci: tur checker reports path
> is down
> Jul 14 03:16:02 ausracdbd01 kernel: device-mapper: multipath: Failing
> path 69:80.
> Jul 14 03:16:02 ausracdbd01 multipathd: sdco: tur checker reports path
> is down
> ....
> </snip>
>
> During this snapshot time, I would like to temporarily turn off the
> verbosity of multipathd. Is this possible? It's mostly an annoyance
> factor for me because our nightly centralized syslog scan picks up words
> like 'fail' and generates a report to the system administrators (which
> we then have to ignore for these messages).
>
Why not just remove the multipath from the multipathd topology before deleting the snapvol ?

multipathd -k"del map mpath19"

Regards,
cvaroqui

--
dm-devel mailing list
dm-devel@redhat.com
https://www.redhat.com/mailman/listinfo/dm-devel

When adding or removing LUNs, I always get multiple multipath messages
reporting that a path is down, sometimes even to devices that I have not
removed. I've tried removing all the SCSI devices (via "echo 1
>/sys/bus/devices/$DEVICE/delete"), flushing the multipath maps,
removing the mpath in question (via "multipathd -k"dep map $mpath"),
prior to unpresenting the LUN or doing a SCSI rescan when presenting a
LUN, all to no avail.

What is proper procedure to completely remove all SCSI devices and
multipath maps so that multipathd will not generate error messages when
the LUN is unpresented from the system?

TIA,

Daniel


> -----Original Message-----
> From: Christophe Varoqui [mailto:christophe.varoqui@free.fr]
> Sent: Saturday, July 19, 2008 3:28 AM
> To: Daniel Keisling
> Cc: device-mapper development
> Subject: Re: Temporarily squelching multipathd errors
>
>
> > Greetings,
> >
> > I perform SAN-based snapshots on OCFS2 and ext3 filesystems under
> > dm-multipath. Before I snapshot the filesystem, I unmount
> and delete
> > the previous night's snapshot LUNs. Doing so causes (correctly)
> > multipathd to output the following errors:
> >
> > <snip>
> > ...
> > Jul 14 03:16:02 ausracdbd01 multipathd: mpath19: remaining
> active paths:
> > 7
> > Jul 14 03:16:02 ausracdbd01 multipathd: sdci: tur checker
> reports path
> > is down
> > Jul 14 03:16:02 ausracdbd01 kernel: device-mapper:
> multipath: Failing
> > path 69:80.
> > Jul 14 03:16:02 ausracdbd01 multipathd: sdco: tur checker
> reports path
> > is down
> > ....
> > </snip>
> >
> > During this snapshot time, I would like to temporarily turn off the
> > verbosity of multipathd. Is this possible? It's mostly an
> annoyance
> > factor for me because our nightly centralized syslog scan
> picks up words
> > like 'fail' and generates a report to the system
> administrators (which
> > we then have to ignore for these messages).
> >
> Why not just remove the multipath from the multipathd
> topology before deleting the snapvol ?
>
> multipathd -k"del map mpath19"
>
> Regards,
> cvaroqui
>
>
>

__________________________________________________ ____________________
This email transmission and any documents, files or previous email
messages attached to it may contain information that is confidential or
legally privileged. If you are not the intended recipient or a person
responsible for delivering this transmission to the intended recipient,
you are hereby notified that you must not read this transmission and
that any disclosure, copying, printing, distribution or use of this
transmission is strictly prohibited. If you have received this transmission
in error, please immediately notify the sender by telephone or return email
and delete the original transmission and its attachments without reading
or saving in any manner.


--
dm-devel mailing list
dm-devel@redhat.com
https://www.redhat.com/mailman/listinfo/dm-devel

> From Daniel Keisling:
>
> What is proper procedure to completely remove all SCSI
> devices and multipath maps so that multipathd will not
> generate error messages when the LUN is unpresented from the system?

This is my recipe:

- Take a note of "multipath -ll" output (SCSI-IDs, long multipath
identifiers, etc)
- Umount filesystem
- Deactivate VG
vgchange -an vg_XXX
- Export VG
vgexport vg_XXX
- Remove multipath
multipath -f <long-multipath-identifier>
- Unpresent/destroy array LUN
- Remove SCSI devices
echo 1 > /sys/bus/scsi/drivers/sd/<SCSI-ID>/delete
... (for all paths)
- Edit fstab


--
Domenico Viggiani

--
dm-devel mailing list
dm-devel@redhat.com
https://www.redhat.com/mailman/listinfo/dm-devel

I am still getting multipath errors when unpresenting and presenting
LUNs. I've traced it down to getting path failures everytime I either
remove or add a LUN to the system. The strange thing is that I get path
errors for maps that aren't related to the LUNs I am deleting or adding.
In this case, I am deleting the LUN with the map 'mpath18,' but other
maps complain whenever the LUN is deleted:

Nov 21 13:47:05 ausracdbd01 multipathd: checker failed path 67:208 in
map orabackup_hs
Nov 21 13:47:05 ausracdbd01 kernel: device-mapper: multipath: Failing
path 67:208.
Nov 21 13:47:08 ausracdbd01 multipathd: checker failed path 70:160 in
map mpath18
Nov 21 13:47:08 ausracdbd01 kernel: device-mapper: multipath: Failing
path 70:160.
Nov 21 13:47:09 ausracdbd01 multipathd: checker failed path 69:0 in map
orabackup_hs
Nov 21 13:47:09 ausracdbd01 kernel: device-mapper: multipath: Failing
path 69:0.
Nov 21 13:47:10 ausracdbd01 multipathd: checker failed path 65:192 in
map ph1d
Nov 21 13:47:10 ausracdbd01 kernel: device-mapper: multipath: Failing
path 65:192.
Nov 21 13:47:10 ausracdbd01 multipathd: checker failed path 65:224 in
map limsd
Nov 21 13:47:10 ausracdbd01 kernel: device-mapper: multipath: Failing
path 65:224.
Nov 21 13:47:10 ausracdbd01 multipathd: checker failed path 65:240 in
map limsd_arch


Sometimes I only get 5-10 "checker failed path" errors, other times I
get 30+.

Can anyone give me any advice? These are OCFS2 filesystems on RHEL5
attached to a HP EVA8000.

[root@ausracdbd01 ~]# rpm -qa | grep mapper
device-mapper-1.02.20-1.el5
device-mapper-1.02.20-1.el5
device-mapper-multipath-0.4.7-12.el5

Relevant /etc/multipath.conf entries:
### HP Storageworks Arrays.
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
bindings_file "/var/lib/multipath/bindings"


devices {

# For EVA3000(HSV101) / EVA5000(HSV111) / EVA4000/6000 / EVA8000
device {
vendor "HP|COMPAQ"
product "HSV1[01]1 (C)COMPAQ|HSV[23][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
}



TIA,

Daniel

> -----Original Message-----
> From: dm-devel-bounces@redhat.com
> [mailto:dm-devel-bounces@redhat.com] On Behalf Of Domenico Viggiani
> Sent: Thursday, October 30, 2008 11:15 AM
> To: 'device-mapper development'
> Subject: [dm-devel] RE: Temporarily squelching multipathd errors
>
> > From Daniel Keisling:
> >
> > What is proper procedure to completely remove all SCSI
> > devices and multipath maps so that multipathd will not
> > generate error messages when the LUN is unpresented from the system?
>
> This is my recipe:
>
> - Take a note of "multipath -ll" output (SCSI-IDs, long multipath
> identifiers, etc)
> - Umount filesystem
> - Deactivate VG
> vgchange -an vg_XXX
> - Export VG
> vgexport vg_XXX
> - Remove multipath
> multipath -f <long-multipath-identifier>
> - Unpresent/destroy array LUN
> - Remove SCSI devices
> echo 1 > /sys/bus/scsi/drivers/sd/<SCSI-ID>/delete
> ... (for all paths)
> - Edit fstab
>
>
> --
> Domenico Viggiani
>
> --
> dm-devel mailing list
> dm-devel@redhat.com
> https://www.redhat.com/mailman/listinfo/dm-devel
>
>

__________________________________________________ ____________________
This email transmission and any documents, files or previous email
messages attached to it may contain information that is confidential or
legally privileged. If you are not the intended recipient or a person
responsible for delivering this transmission to the intended recipient,
you are hereby notified that you must not read this transmission and
that any disclosure, copying, printing, distribution or use of this
transmission is strictly prohibited. If you have received this transmission
in error, please immediately notify the sender by telephone or return email
and delete the original transmission and its attachments without reading
or saving in any manner.


--
dm-devel mailing list
dm-devel@redhat.com
https://www.redhat.com/mailman/listinfo/dm-devel

On Fri, Nov 21, 2008 at 01:54:01PM -0600, Daniel Keisling wrote:
> I am still getting multipath errors when unpresenting and presenting
> LUNs. I've traced it down to getting path failures everytime I either
> remove or add a LUN to the system. The strange thing is that I get path
> errors for maps that aren't related to the LUNs I am deleting or adding.
> In this case, I am deleting the LUN with the map 'mpath18,' but other
> maps complain whenever the LUN is deleted:
>
> Nov 21 13:47:05 ausracdbd01 multipathd: checker failed path 67:208 in
> map orabackup_hs
> Nov 21 13:47:05 ausracdbd01 kernel: device-mapper: multipath: Failing
> path 67:208.
> Nov 21 13:47:08 ausracdbd01 multipathd: checker failed path 70:160 in
> map mpath18
> Nov 21 13:47:08 ausracdbd01 kernel: device-mapper: multipath: Failing
> path 70:160.
> Nov 21 13:47:09 ausracdbd01 multipathd: checker failed path 69:0 in map
> orabackup_hs
> Nov 21 13:47:09 ausracdbd01 kernel: device-mapper: multipath: Failing
> path 69:0.
> Nov 21 13:47:10 ausracdbd01 multipathd: checker failed path 65:192 in
> map ph1d
> Nov 21 13:47:10 ausracdbd01 kernel: device-mapper: multipath: Failing
> path 65:192.
> Nov 21 13:47:10 ausracdbd01 multipathd: checker failed path 65:224 in
> map limsd
> Nov 21 13:47:10 ausracdbd01 kernel: device-mapper: multipath: Failing
> path 65:224.
> Nov 21 13:47:10 ausracdbd01 multipathd: checker failed path 65:240 in
> map limsd_arch
>

Can you attach the 'multipath -ll' output? Run the program before
you delete/add the LUNs and after, please.

How do you delete/add the LUNs? Are you re-using the LUN number, or are you
actually deleting a LUN, deleting the block device as well, and then
when adding (with a new LUN number higher than the deleted one), rescanning
the SCSI host channel to get the new block device for the LUN?

--
dm-devel mailing list
dm-devel@redhat.com
https://www.redhat.com/mailman/listinfo/dm-devel