[Rear-users] [ rear-Bugs-3441168 ] BUG BUG BUG! Could not determine size of disk

Domin, Pavol pavol.domin at hp.com
Thu Nov 24 09:21:11 CET 2011


Hi,

We have many systems that boot from partitioned multipathed lun (in the mdraid)

Layout (on SLES11) looks like:

# multipath -l 3600508b4000ad7f5000070000adf0000
3600508b4000ad7f5000070000adf0000 dm-2 HP,HSV210
size=50G features='1 queue_if_no_path' hwhandler='1 hp_sw' wp=rw
|-+- policy='round-robin 0' prio=0 status=active
| |- 0:0:6:1  sdy  65:128 active undef running
| |- 1:0:4:1  sdaw 67:0   active undef running
| |- 1:0:6:1  sdbe 67:128 active undef running
| |- 2:0:5:1  sdcg 69:64  active undef running
| |- 2:0:6:1  sdck 69:128 active undef running
| |- 3:0:4:1  sddi 71:0   active undef running
| |- 3:0:6:1  sddq 71:128 active undef running
| `- 0:0:4:1  sdq  65:0   active undef running
`-+- policy='round-robin 0' prio=0 status=enabled
  |- 0:0:5:1  sdu  65:64  active undef running
  |- 1:0:5:1  sdba 67:64  active undef running
  |- 1:0:7:1  sdbi 67:192 active undef running
  |- 2:0:4:1  sdcc 69:0   active undef running
  |- 2:0:7:1  sdco 69:192 active undef running
  |- 3:0:5:1  sddm 71:64  active undef running
  |- 3:0:7:1  sddu 71:192 active undef running
  `- 0:0:7:1  sdac 65:192 active undef running

# dmsetup table | grep 3600508b4000ad7f5000070000adf0000
3600508b4000ad7f5000070000adf0000: 0 104857600 multipath 1 queue_if_no_path 1 hp_sw 2 1 round-robin 0 8 1 65:128 1000 67:0 1000 67:128 1000 69:64 1000 69:128 1000 71:0 1000 71:128 1000 65:0 1000 round-robin 0 8 1 65:64 1000 67:64 1000 67:192 1000 69:0 1000 69:192 1000 71:64 1000 71:192 1000 65:192 1000
3600508b4000ad7f5000070000adf0000_part2: 0 83666520 linear 253:2 208845
3600508b4000ad7f5000070000adf0000_part1: 0 208782 linear 253:2 63

# ls /sys/block/sdy
alignment_offset  capability  device     holders   power  range      ro    sdy2  slaves  subsystem
bdi               dev         ext_range  inflight  queue  removable  sdy1  size  stat    uevent

# ls /sys/block/sdy/sdy[12]
/sys/block/sdy/sdy1:
alignment_offset  dev  holders  inflight  partition  power  size  start  stat  subsystem  uevent

/sys/block/sdy/sdy2:
alignment_offset  dev  holders  inflight  partition  power  size  start  stat  subsystem  uevent


Kind regards

Pavol



> -----Original Message-----
> From: Jeroen Hoekx [mailto:jeroen.hoekx at hamok.be]
> Sent: Wednesday, November 23, 2011 9:49 PM
> To: Discussions about Relax & Recover
> Subject: Re: [Rear-users] [ rear-Bugs-3441168 ] BUG BUG BUG! Could not
> determine size of disk
> 
> On 23 November 2011 20:46, Dag Wieers <dag at wieers.com> wrote:
> >
> > Ok :) A heads-up would have been nice. Not sure if Jeroen wanted his
> > latest patches included in 1.12.0, or whether he was waiting for the
> > branch in order to commit ;-) Let's assume the latter.
> 
> Waiting for the branch... The work's not finished yet, but this should at least fix
> this particular issue.
> 
> I would also like to solve the issue in this bug report for the old code, as my last
> commits have a larger potential bug surface. I'm now waiting for a reply from
> the reporter as I do not have any multipath enabled systems that I can check
> these days. None of my previous multipath systems which I used to develop and
> test had partitions on the multipath devices, so this went unnoticed. I really
> should create a VM that uses iSCSI devices from another one...
> 
> It's not a regression from 1.11 to 1.12, though, but from 1.10, so it's not like I
> broke something very recently :-)
> 
> If anyone can check the sysfs layout of multipath devices with partitions, that
> would be appreciated:
> # ls /sys/block/sdX
> # ls /sys/block/sdX/sdX1 (if it exists)
> 
> Greetings,
> 
> Jeroen
> 
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure contains a
> definitive record of customers, application performance, security threats,
> fraudulent activity, and more. Splunk takes this data and makes sense of it. IT
> sense. And common sense.
> http://p.sf.net/sfu/splunk-novd2d
> _______________________________________________
> Rear-users mailing list
> Rear-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/rear-users




More information about the rear-users mailing list