* unmapping luns doesn't work (good)
@ 2010-10-18 15:24 Rudy Gevaert
2010-10-21 7:28 ` Chauhan, Vijay
0 siblings, 1 reply; 5+ messages in thread
From: Rudy Gevaert @ 2010-10-18 15:24 UTC (permalink / raw)
To: dm-devel
Hello list,
This is my first real experience with device mapper and multipath-tools.
I have several luns exported on an IBM 3400 to one host.
I have several problems when I unmap a lun:
1) If I don't remove the multipath (before I unmap) I can not remove (by
force too) the multipath for that lun. (Note the lun is not mounted.)
2) Removing the multipath and then unmapping the lun works, but then
doing a rescan of the bus **sometimes** gives me a new lun with name '1'.
Case 1:
-------
multipath -l
mail17r-ds3400-1 (3600a0b80005a67b6000004764cac4912) dm-2 IBM,1726-4xx
FAStT
size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=rw
|-+- policy='round-robin 0' prio=-1 status=active
| `- 0:0:0:6 sde 8:64 active undef running
`-+- policy='round-robin 0' prio=-1 status=enabled
`- 5:0:0:6 sdj 8:144 active undef running
root@cyrprd1:~# dmsetup table
mail17r-ds3400-1: 0 1153433600 multipath 1 queue_if_no_path 1 rdac 2 1
round-robin 0 1 1 8:64 1000 round-robin 0 1 1 8:144 1000
I now remove mapping through the storage software.
root@cyrprd1:~# multipath -f mail14r-ds3400-1
Oct 18 17:07:09 | mail14r-ds3400-1: map in use
root@cyrprd1:~# multipath -l mail14r-ds3400-1
mail14r-ds3400-1 (3600a0b80005a591a000004294cac468b) dm-0 IBM,1726-4xx
FAStT
size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=rw
|-+- policy='round-robin 0' prio=-1 status=enabled
| `- 5:0:0:2 sdg 8:96 failed undef running
`-+- policy='round-robin 0' prio=-1 status=enabled
`- 0:0:0:2 sdb 8:16 failed undef running
root@cyrprd1:~#
Case 2:
--------
Sorry couldn't reproduce at the moment...
I'm running
kernel 2.6.32-5
multipath-tools 0.4.9-1
os: Debian Squeeze
Thanks!
Rudy
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: unmapping luns doesn't work (good)
2010-10-18 15:24 unmapping luns doesn't work (good) Rudy Gevaert
@ 2010-10-21 7:28 ` Chauhan, Vijay
2010-10-21 14:52 ` Rudy Gevaert
0 siblings, 1 reply; 5+ messages in thread
From: Chauhan, Vijay @ 2010-10-21 7:28 UTC (permalink / raw)
To: device-mapper development
On Monday, October 18, 2010 8:54 PM, Rudy Gevaert wrote:
> Case 1:
> -------
> multipath -l
>
> mail17r-ds3400-1 (3600a0b80005a67b6000004764cac4912) dm-2
> IBM,1726-4xx
> FAStT
> size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=rw
> |-+- policy='round-robin 0' prio=-1 status=active
> | `- 0:0:0:6 sde 8:64 active undef running
> `-+- policy='round-robin 0' prio=-1 status=enabled
> `- 5:0:0:6 sdj 8:144 active undef running
>
> root@cyrprd1:~# dmsetup table
> mail17r-ds3400-1: 0 1153433600 multipath 1 queue_if_no_path 1
> rdac 2 1
> round-robin 0 1 1 8:64 1000 round-robin 0 1 1 8:144 1000
>
> I now remove mapping through the storage software.
Try with disabling the queuing before removing the map.
dmsetup message mail14r-ds3400-1 0 "fail_if_no_path"
>
> root@cyrprd1:~# multipath -f mail14r-ds3400-1
> Oct 18 17:07:09 | mail14r-ds3400-1: map in use
>
Regards,
Vijay
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: unmapping luns doesn't work (good)
2010-10-21 7:28 ` Chauhan, Vijay
@ 2010-10-21 14:52 ` Rudy Gevaert
2010-10-22 8:35 ` Chauhan, Vijay
0 siblings, 1 reply; 5+ messages in thread
From: Rudy Gevaert @ 2010-10-21 14:52 UTC (permalink / raw)
To: dm-devel
On 10/21/2010 09:28 AM, Chauhan, Vijay wrote:
> Try with disabling the queuing before removing the map.
>
> dmsetup message mail14r-ds3400-1 0 "fail_if_no_path"
Hi
When I do that, and then unmap the lun.
If I then rescan an issue multipath I get:
root@cyrprd1:~# multipath
create: 1 undef IBM,1726-4xx FAStT
size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=undef
`-+- policy='round-robin 0' prio=0 status=undef
|- 0:0:0:3 sdd 8:48 undef faulty running
`- 5:0:0:3 sdl 8:176 undef faulty running
<hang>
Any ideas?
Thanks!
Rudy
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: unmapping luns doesn't work (good)
2010-10-21 14:52 ` Rudy Gevaert
@ 2010-10-22 8:35 ` Chauhan, Vijay
2010-10-27 13:26 ` Rudy Gevaert
0 siblings, 1 reply; 5+ messages in thread
From: Chauhan, Vijay @ 2010-10-22 8:35 UTC (permalink / raw)
To: device-mapper development
On Thur, October 21, 2010 8:22 PM, Rudy Gevaert wrote:
> On 10/21/2010 09:28 AM, Chauhan, Vijay wrote:
>
> > Try with disabling the queuing before removing the map.
> >
> > dmsetup message mail14r-ds3400-1 0 "fail_if_no_path"
>
> Hi
>
> When I do that, and then unmap the lun.
>
> If I then rescan an issue multipath I get:
>
> root@cyrprd1:~# multipath
> create: 1 undef IBM,1726-4xx FAStT
> size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=undef
> `-+- policy='round-robin 0' prio=0 status=undef
> |- 0:0:0:3 sdd 8:48 undef faulty running
> `- 5:0:0:3 sdl 8:176 undef faulty running
> <hang>
>
>
You may not get into this issue if you follow these steps:
step 1) Delete the paths for the dm device to be unmapped. (Ensure that this dm device is not being used for IO)
Once the last path is deleted, multipathd will remove this map.
step 2) Unmap the LUN from storage.
Hope this helps.
Regards,
Vijay
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: unmapping luns doesn't work (good)
2010-10-22 8:35 ` Chauhan, Vijay
@ 2010-10-27 13:26 ` Rudy Gevaert
0 siblings, 0 replies; 5+ messages in thread
From: Rudy Gevaert @ 2010-10-27 13:26 UTC (permalink / raw)
To: dm-devel
Hello Vijay,
On 10/22/2010 10:35 AM, Chauhan, Vijay wrote:
>
> On Thur, October 21, 2010 8:22 PM, Rudy Gevaert wrote:
>> On 10/21/2010 09:28 AM, Chauhan, Vijay wrote:
>>
>>> Try with disabling the queuing before removing the map.
>>>
>>> dmsetup message mail14r-ds3400-1 0 "fail_if_no_path"
>>
>> Hi
>>
>> When I do that, and then unmap the lun.
>>
>> If I then rescan an issue multipath I get:
>>
>> root@cyrprd1:~# multipath
>> create: 1 undef IBM,1726-4xx FAStT
>> size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=undef
>> `-+- policy='round-robin 0' prio=0 status=undef
>> |- 0:0:0:3 sdd 8:48 undef faulty running
>> `- 5:0:0:3 sdl 8:176 undef faulty running
>> <hang>
>>
>>
>
> You may not get into this issue if you follow these steps:
>
> step 1) Delete the paths for the dm device to be unmapped. (Ensure that this dm device is not being used for IO)
> Once the last path is deleted, multipathd will remove this map.
>
> step 2) Unmap the LUN from storage.
>
> Hope this helps.
>
> Regards,
> Vijay
Unfortunately I then get up with:
root@cyrprd3:~# multipath -l
1 dm-3 IBM,1726-4xx FAStT
size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=rw
`-+- policy='round-robin 0' prio=-30 status=enabled
|- 5:0:0:10 sdb 8:16 failed undef running
|- 6:0:0:10 sdc 8:32 failed undef running
|- 5:0:0:12 sdd 8:48 failed undef running
|- 6:0:0:12 sde 8:64 failed undef running
|- 5:0:0:20 sdf 8:80 failed undef running
|- 6:0:0:20 sdg 8:96 failed undef running
|- 5:0:0:9 sdh 8:112 failed undef running
|- 6:0:0:9 sdi 8:128 failed undef running
|- 5:0:0:11 sdj 8:144 failed undef running
|- 6:0:0:11 sdk 8:160 failed undef running
|- 5:0:0:13 sdl 8:176 failed undef running
|- 6:0:0:13 sdm 8:192 failed undef running
|- 5:0:0:8 sdn 8:208 failed undef running
|- 6:0:0:8 sdo 8:224 failed undef running
|- 5:0:0:21 sdp 8:240 failed undef running
|- 6:0:0:21 sdq 65:0 failed undef running
|- 5:0:0:19 sdr 65:16 failed undef running
|- 6:0:0:19 sds 65:32 failed undef running
|- 5:0:0:18 sdt 65:48 failed undef running
|- 6:0:0:18 sdu 65:64 failed undef running
|- 5:0:0:17 sdv 65:80 failed undef running
|- 6:0:0:17 sdw 65:96 failed undef running
|- 5:0:0:22 sdx 65:112 failed undef running
|- 6:0:0:22 sdy 65:128 failed undef running
|- 5:0:0:16 sdz 65:144 failed undef running
|- 6:0:0:16 sdaa 65:160 failed undef running
|- 5:0:0:15 sdab 65:176 failed undef running
|- 6:0:0:15 sdac 65:192 failed undef running
|- 5:0:0:14 sdad 65:208 failed undef running
`- #:#:#:# - #:# failed undef running
(for the last line, I also echo'd 1 to the remove file. Doing for all
devices gets me:
1 dm-3 ,
size=550G features='1 queue_if_no_path' hwhandler='1 rdac' wp=rw
`-+- policy='round-robin 0' prio=-30 status=active
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# active undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
|- #:#:#:# - #:# failed undef running
`- #:#:#:# - #:# failed undef running
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2010-10-27 13:26 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-10-18 15:24 unmapping luns doesn't work (good) Rudy Gevaert
2010-10-21 7:28 ` Chauhan, Vijay
2010-10-21 14:52 ` Rudy Gevaert
2010-10-22 8:35 ` Chauhan, Vijay
2010-10-27 13:26 ` Rudy Gevaert
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.