About using multipath in SLES 10

About using multipath in SLES 10

[Erez Zilber]

Hi,

I guess that I'm missing something:

I started /etc/init.d/multipathd on SLES 10 sp1 and now I see the
following devices:

seed1:~ # multipath -ll
SATA_HDS728080PLA380_PFDB32E8TPKR5Mdm-0 ATA,HDS728080PLA380
[size=77G][features=0][hwhandler=0]
\_ round-robin 0 [prio=1][active]
 \_ 3:0:0:0  sdb 8:16  [active][ready]
3600c0ff000d508d7def2894702000000dm-1 DotHill,R/Evo 2730-2R
[size=82G][features=0][hwhandler=0]
\_ round-robin 0 [prio=1][active]
 \_ 16:0:0:0 sdc 8:32  [active][ready]
\_ round-robin 0 [prio=1][enabled]
 \_ 17:0:0:0 sdd 8:48  [active][ready]
3600c0ff000d508d7dff2894701000000dm-2 DotHill,R/Evo 2730-2R
[size=82G][features=0][hwhandler=0]
\_ round-robin 0 [prio=1][active]
 \_ 18:0:0:0 sde 8:64  [active][ready]
\_ round-robin 0 [prio=1][enabled]
 \_ 19:0:0:0 sdf 8:80  [active][ready]

I understand that sdc & sdd are mapped to dm-1. However, I don't
understand how to use dm-1. For example, I can't run INQUIRY on it:

seed1:~ # sg_inq /dev/dm-1
Both SCSI INQUIRY and fetching ATA information failed on /dev/dm-1

I guess that I'm doing something wrong here... What should I do?

Thanks,
Erez

Re: About using multipath in SLES 10

[malahal]

Erez Zilber [erezz@voltaire.com] wrote:
> Hi,
> 
> I guess that I'm missing something:
> 
> I started /etc/init.d/multipathd on SLES 10 sp1 and now I see the
> following devices:
> 
> 3600c0ff000d508d7def2894702000000dm-1 DotHill,R/Evo 2730-2R
> [size=82G][features=0][hwhandler=0]
> \_ round-robin 0 [prio=1][active]
>  \_ 16:0:0:0 sdc 8:32  [active][ready]
> \_ round-robin 0 [prio=1][enabled]
>  \_ 17:0:0:0 sdd 8:48  [active][ready]
> 
> I understand that sdc & sdd are mapped to dm-1. However, I don't
> understand how to use dm-1. For example, I can't run INQUIRY on it:
> 
> seed1:~ # sg_inq /dev/dm-1
> Both SCSI INQUIRY and fetching ATA information failed on /dev/dm-1

All things that work on /dev/sdc may not work on /dev/dm-1. You should
use /dev/sdc instead of /dev/dm-1 for those cases. You should be able to
use /dev/dm-1 for normal I/O (like dd, mounting etc.)

Thanks, Malahal.

RE: About using multipath in SLES 10

[Erez Zilber]

[-- Attachment #1: Type: text/plain, Size: 1336 bytes --]

> > I understand that sdc & sdd are mapped to dm-1. However, I don't
> > understand how to use dm-1. For example, I can't run INQUIRY on it:
> >
> > seed1:~ # sg_inq /dev/dm-1
> > Both SCSI INQUIRY and fetching ATA information failed on /dev/dm-1
> 
> All things that work on /dev/sdc may not work on /dev/dm-1. You should
> use /dev/sdc instead of /dev/dm-1 for those cases. You should be able to
> use /dev/dm-1 for normal I/O (like dd, mounting etc.)

I understand. I've tried disktest, and it work ok. I have another question - I'm running a single iSCSI initiator against 2 iSCSI target. I started disktest on /dev/dm-1 and after a few seconds, I disconnected the primary target. It took ~2 minutes until it failed over to the secondary device. During these ~2 minutes, it seemed that disktest is still reading data from the target, just slower. That's very strange. After failover was completed, I was able to use the secondary device.
 
I have 2 questions:

1.	
	Which parameter in  multipath.conf do I need to change in order to failover in a few seconds? Is it polling_interval? I saw that the default value is 5 seconds, which should be ok.
2.	
	Before multipath decided that it needs to failover, why did I see that traffic is still running? It had no device to work with at that time.

Thanks,

Erez


[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 4152 bytes --]

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

RE: About using multipath in SLES 10

[Erez Zilber]

[-- Attachment #1: Type: text/plain, Size: 579 bytes --]

And another question:
 
I've added the following entry to multipath.conf:
 
        multipath {
                wwid    3600c0ff000d508d7def2894702000000
                alias   red
        }
 
which should represent the following device:
 
3600c0ff000d508d7def2894702000000dm-1 DotHill,R/Evo 2730-2R
[size=82G][features=0][hwhandler=0]
\_ round-robin 0 [prio=0][enabled]
 \_ 16:0:0:0 sdc 8:32  [failed][faulty]
\_ round-robin 0 [prio=0][active]
 \_ 17:0:0:0 sdd 8:48  [active][undef] 
 
Where should I see "red"? Is it under /dev somewhere?
 
Thanks,
Erez

[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 3860 bytes --]

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: About using multipath in SLES 10

[malahal]

Erez Zilber [erezz@voltaire.com] wrote:
> And another question:
>  
> I've added the following entry to multipath.conf:
>  
>         multipath {
>                 wwid    3600c0ff000d508d7def2894702000000
>                 alias   red
>         }
>  
> which should represent the following device:
>  
> 3600c0ff000d508d7def2894702000000dm-1 DotHill,R/Evo 2730-2R
> [size=82G][features=0][hwhandler=0]
> \_ round-robin 0 [prio=0][enabled]
>  \_ 16:0:0:0 sdc 8:32  [failed][faulty]
> \_ round-robin 0 [prio=0][active]
>  \_ 17:0:0:0 sdd 8:48  [active][undef] 
>  
> Where should I see "red"? Is it under /dev somewhere?

If it worked, it should be /dev/mapper/red, but I have a suspicion that it
didn't work for you based on your above 'multipath -l' output.

Thanks, Malahal.

multipath device name

[Erez Zilber]

[-- Attachment #1: Type: text/plain, Size: 913 bytes --]

> Erez Zilber [erezz@voltaire.com] wrote:
> > And another question:
> > 
> > I've added the following entry to multipath.conf:
> > 
> >         multipath {
> >                 wwid    3600c0ff000d508d7def2894702000000
> >                 alias   red
> >         }
> > 
> > which should represent the following device:
> > 
> > 3600c0ff000d508d7def2894702000000dm-1 DotHill,R/Evo 2730-2R
> > [size=82G][features=0][hwhandler=0]
> > \_ round-robin 0 [prio=0][enabled]
> >  \_ 16:0:0:0 sdc 8:32  [failed][faulty]
> > \_ round-robin 0 [prio=0][active]
> >  \_ 17:0:0:0 sdd 8:48  [active][undef]
> > 
> > Where should I see "red"? Is it under /dev somewhere?
> 
> If it worked, it should be /dev/mapper/red, but I have a suspicion that it
> didn't work for you based on your above 'multipath -l' output.

No, it didn't work. Do I need to do anything special to make it work?
 
Thanks,
Erez

[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 4222 bytes --]

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: multipath device name

[malahal]

Erez Zilber [erezz@voltaire.com] wrote:
> > Erez Zilber [erezz@voltaire.com] wrote:
> > > And another question:
> > > 
> > > I've added the following entry to multipath.conf:
> > > 
> > >         multipath {
> > >                 wwid    3600c0ff000d508d7def2894702000000
> > >                 alias   red
> > >         }
> > > 
> > > which should represent the following device:
> > > 
> > > 3600c0ff000d508d7def2894702000000dm-1 DotHill,R/Evo 2730-2R
> > > [size=82G][features=0][hwhandler=0]
> > > \_ round-robin 0 [prio=0][enabled]
> > >  \_ 16:0:0:0 sdc 8:32  [failed][faulty]
> > > \_ round-robin 0 [prio=0][active]
> > >  \_ 17:0:0:0 sdd 8:48  [active][undef]
> > > 
> > > Where should I see "red"? Is it under /dev somewhere?
> > 
> > If it worked, it should be /dev/mapper/red, but I have a suspicion that it
> > didn't work for you based on your above 'multipath -l' output.
> 
> No, it didn't work. Do I need to do anything special to make it work?

Did you remove the old device and re-run 'multipath' after making that
change to multipath.conf?

RE: multipath device name

[Erez Zilber]

[-- Attachment #1: Type: text/plain, Size: 1166 bytes --]

> > > If it worked, it should be /dev/mapper/red, but I have a suspicion that it
> > > didn't work for you based on your above 'multipath -l' output.
> >
> > No, it didn't work. Do I need to do anything special to make it work?
> 
> Did you remove the old device and re-run 'multipath' after making that
> change to multipath.conf?

Do you mean 'multipath -F'? I did that and then ran ' multipath -p failover'.
 
All I see is this:
 
seed1:~ # ll /dev/mapper/
total 0
brw------- 1 root root 253, 1 2008-01-28 21:53 3600c0ff000d508d7def2894702000000
brw------- 1 root root 253, 2 2008-01-28 21:53 3600c0ff000d508d7dff2894701000000
lrwxrwxrwx 1 root root     16 2008-01-28 11:10 control -> ../device-mapper
brw------- 1 root root 253, 0 2008-01-28 21:53 SATA_HDS728080PLA380_PFDB32E8TPKR5M
brw------- 1 root root 253, 3 2008-01-28 21:53 SATA_HDS728080PLA380_PFDB32E8TPKR5M-part1
brw------- 1 root root 253, 4 2008-01-28 21:53 SATA_HDS728080PLA380_PFDB32E8TPKR5M-part2
brw------- 1 root root 253, 5 2008-01-28 21:53 SATA_HDS728080PLA380_PFDB32E8TPKR5M-part3
brw------- 1 root root 253, 6 2008-01-28 21:53 SATA_HDS728080PLA380_PFDB32E8TPKR5M-part4

[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 3809 bytes --]

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: multipath device name

[malahal]

Erez Zilber [erezz@voltaire.com] wrote:
> > > > If it worked, it should be /dev/mapper/red, but I have a suspicion that it
> > > > didn't work for you based on your above 'multipath -l' output.
> > >
> > > No, it didn't work. Do I need to do anything special to make it work?
> > 
> > Did you remove the old device and re-run 'multipath' after making that
> > change to multipath.conf?
> 
> Do you mean 'multipath -F'? I did that and then ran ' multipath -p failover'.

Yes, I meant the above. Check if you have any problems with your
multipath.conf file. Make sure you enclose "multipath {}" within
"multipaths {}" block. Run 'multipath -t' and make sure your changes are
listed in its output.

What is your multipath.conf file?

RE: multipath device name

[Erez Zilber]

[-- Attachment #1: Type: text/plain, Size: 721 bytes --]

> Yes, I meant the above. Check if you have any problems with your
> multipath.conf file. Make sure you enclose "multipath {}" within
> "multipaths {}" block. Run 'multipath -t' and make sure your changes are
> listed in its output.
> 
> What is your multipath.conf file?

Problem solved:
I copied /usr/share/doc/packages/multipath-tools/multipath.conf.annotated and uncommented the block with the "red" device. The problem is that I didn't remove the comments from 'multipaths {' and the '}' at the end of the block. After uncommenting it, I see /dev/mapper/red.
 
Any idea about the other question that I had on this thread (about slow and strange fail over)? Any help will be appreciated.
 
Thanks,
Erez

[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 3813 bytes --]

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

RE: multipath device name

[Erez Zilber]

[-- Attachment #1: Type: text/plain, Size: 274 bytes --]

> What is your multipath.conf file?

I'm using a conf file that looks like this:
 
defaults {
        user_friendly_names yes
}

multipaths {
        multipath {
                wwid    3600c0ff000d508d7def2894702000000
                alias   red
        }
}

[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 3801 bytes --]

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: About using multipath in SLES 10

[malahal]

Erez Zilber [erezz@voltaire.com] wrote:
> I understand. I've tried disktest, and it work ok. I have another
> question - I'm running a single iSCSI initiator against 2 iSCSI
> target. I started disktest on /dev/dm-1 and after a few seconds, I
> disconnected the primary target. It took ~2 minutes until it failed
> over to the secondary device. During these ~2 minutes, it seemed that
> disktest is still reading data from the target, just slower. That's
> very strange. After failover was completed, I was able to use the
> secondary device.

You said, "it seemed that disktest is still reading data from the
target". Maybe it is reading some cached data (read-ahead or some other
such thing...) Also, the multipath kernel module would not know that you
have disconnected the target until an I/O fails. Depending on your error
injection and the susystems design, the I/O failure could be as a result
of a timeout (this depends on your/distro setting, generally 1 minute).

> I have 2 questions:
> 
> 1.	> 	Which parameter in  multipath.conf do I need to change
> in order to failover in a few seconds? Is it polling_interval? I saw
> that the default value is 5 seconds, which should be ok.

Not an expert, but polling interval can't change the failover time. It
may change the 'failback' time though!

> 2.	Before multipath decided that it needs to failover, why did I
> see that traffic is still running? It had no device to work with at
> that time.

Depends on 'how you saw the traffic'! Could be false alarm.

RE: multipath device name

[Ariel Gottlieb]

[-- Attachment #1.1: Type: text/plain, Size: 1273 bytes --]

Erez, 

Add the following to your multipath.conf  file:

 

defaults {

        udev_dir                /dev

        polling_interval        2

        failback                immediate

        no_path_retry           3

}

 

Regards, 

Ariel 

________________________________

From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com]
On Behalf Of Erez Zilber
Sent: Monday, January 28, 2008 10:23 PM
To: device-mapper development
Subject: RE: [dm-devel] multipath device name

 

> Yes, I meant the above. Check if you have any problems with your
> multipath.conf file. Make sure you enclose "multipath {}" within
> "multipaths {}" block. Run 'multipath -t' and make sure your changes
are
> listed in its output.
> 
> What is your multipath.conf file?

Problem solved:

I copied
/usr/share/doc/packages/multipath-tools/multipath.conf.annotated and
uncommented the block with the "red" device. The problem is that I
didn't remove the comments from 'multipaths {' and the '}' at the end of
the block. After uncommenting it, I see /dev/mapper/red.

 

Any idea about the other question that I had on this thread (about slow
and strange fail over)? Any help will be appreciated.

 

Thanks,

Erez


[-- Attachment #1.2: Type: text/html, Size: 7789 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: multipath device name

[Erez Zilber]

malahal@us.ibm.com wrote:

> Erez Zilber [erezz@voltaire.com] wrote:
>   
>>>>> If it worked, it should be /dev/mapper/red, but I have a suspicion that it
>>>>> didn't work for you based on your above 'multipath -l' output.
>>>>>           
>>>> No, it didn't work. Do I need to do anything special to make it work?
>>>>         
>>> Did you remove the old device and re-run 'multipath' after making that
>>> change to multipath.conf?
>>>       
>> Do you mean 'multipath -F'? I did that and then ran ' multipath -p failover'.
>>     
>
> Yes, I meant the above. Check if you have any problems with your
> multipath.conf file. Make sure you enclose "multipath {}" within
> "multipaths {}" block. Run 'multipath -t' and make sure your changes are
> listed in its output.
>
> What is your multipath.conf file?
>
> --
> dm-devel mailing list
> dm-devel@redhat.com
> https://www.redhat.com/mailman/listinfo/dm-devel
>   

I made some changes in the conf file and some more changes in the iSCSI
initiator conf file, and it work ok now.

Thanks,
Erez