Re install MSA30

[Fabian Crea]

I have a MSA30 with 5 disk of 146G each one, this is in first array

I have a MSA30 connected to a server RP3410, the storage has 5 disk of 146GB each one in the first 5 bays, the server has two controllers scsii connected to storage, all the discs seen by a single channel, the idea is to optimize the reading and to move two discs so that takes the other controller from storage, since I do this new reoganizacion, somebody can help me, please, thank you very much.

[Torsten.]

Hi,

which connector module is installed in the MSA30? 1, 2 or 4 connectors?

Are you using mirroring (LVM) or a smart array card?

Please post an "ioscan -fn" if possible.

[Fabian Crea]

Hi,

only 2 connectors,
I have LVM

The attach is ioscan.

thanks
FC

[Fabian Crea]

Sorry, I have 4 connectors. Only 2 used.
THZ

[Torsten.]

Looks like a rp3440 to me.

You have several options.
(Be aware that modifications will impact your LVM configuration).

First option is to connect the same bus in the MSA30 to the other SCSI HBA (slot 2) in your server.

In this case you have to modify the initiator ID before connecting the cable.

mptconfig -i 6 /dev/mpt6
mptconfig -i 6 /dev/mpt7

(for the "spare" card in slot 2)

Now you have redundant pathes to the disks and may add the "new" pathes to your volume group config.


The other option is to move 2 disks from the left to the right half of the chassis, cable the same way and mirror across the busses.
This will bring more performance, because 2 seperate busses are used.

[Guillermo Rossney]

Thanks Torsten!
It seeams as my partner and I tried to solve our problem in the same way!

I'll continue on this. Sorry by the double effort.

Our primary target is performance.

As i said on the other thread (thanks and sorry, again), our idea was move 2 disks to the other half of the chasis. But

- Our provider said that is not possible.
- We need to know the impact of that change on data already transferred to disks and actual LVM configuration.


Thanks a lot!

[Fabian Crea]

Thank very much for all, we will working in this case.

[Torsten.]

Please post the output of "vgdisplay -v", this will help to know about the impact on changing the disk locations.

(Please close the other thread)

[Fabian Crea]

Thanks
the output:

# vgdisplay -v
--- Volume groups ---
VG Name /dev/vg00
VG Write Access read/write
VG Status available
Max LV 255
Cur LV 10
Open LV 10
Max PV 16
Cur PV 3
Act PV 3
Max PE per PV 4384
VGDA 6
PE Size (Mbytes) 32
Total PE 8748
Alloc PE 3096
Free PE 5652
Total PVG 0
Total Spare PVs 1
Total Spare PVs in use 0

--- Logical volumes ---
LV Name /dev/vg00/lvol1
LV Status available/syncd
LV Size (Mbytes) 320
Current LE 10
Allocated PE 20
Used PV 2

LV Name /dev/vg00/lvol2
LV Status available/syncd
LV Size (Mbytes) 4096
Current LE 128
Allocated PE 256
Used PV 2

LV Name /dev/vg00/lvol3
LV Status available/syncd
LV Size (Mbytes) 256
Current LE 8
Allocated PE 16
Used PV 2

LV Name /dev/vg00/lvol4
LV Status available/syncd
LV Size (Mbytes) 2048
Current LE 64
Allocated PE 128
Used PV 2

LV Name /dev/vg00/lvol5
LV Status available/syncd
LV Size (Mbytes) 2048
Current LE 64
Allocated PE 128
Used PV 2

LV Name /dev/vg00/lvol6
LV Status available/syncd
LV Size (Mbytes) 3424
Current LE 107
Allocated PE 214
Used PV 2

LV Name /dev/vg00/lvol7
LV Status available/syncd
LV Size (Mbytes) 2496
Current LE 78
Allocated PE 156
Used PV 2

LV Name /dev/vg00/lvol8
LV Status available/syncd
LV Size (Mbytes) 4608
Current LE 144
Allocated PE 288
Used PV 2

LV Name /dev/vg00/lvu11
LV Status available/syncd
LV Size (Mbytes) 30016
Current LE 938
Allocated PE 1876
Used PV 2

LV Name /dev/vg00/lvsql
LV Status available/syncd
LV Size (Mbytes) 224
Current LE 7
Allocated PE 14
Used PV 2


--- Physical volumes ---
PV Name /dev/dsk/c2t0d0
PV Status available
Total PE 4374
Free PE 2826
Autoswitch On

PV Name /dev/dsk/c2t1d0
PV Status available
Total PE 4374
Free PE 2826
Autoswitch On

PV Name /dev/dsk/c3t2d0
PV Status available/standby spare
Total PE 4374
Free PE 4374
Autoswitch On


VG Name /dev/vgprd
VG Write Access read/write
VG Status available
Max LV 255
Cur LV 7
Open LV 7
Max PV 16
Cur PV 5
Act PV 5
Max PE per PV 4375
VGDA 10
PE Size (Mbytes) 32
Total PE 21870
Alloc PE 20160
Free PE 1710
Total PVG 0
Total Spare PVs 0
Total Spare PVs in use 0

--- Logical volumes ---
LV Name /dev/vgprd/lvsql
LV Status available/syncd
LV Size (Mbytes) 1024
Current LE 32
Allocated PE 64
Used PV 2

LV Name /dev/vgprd/lvu12
LV Status available/syncd
LV Size (Mbytes) 133120
Current LE 4160
Allocated PE 8320
Used PV 2

LV Name /dev/vgprd/lvu13
LV Status available/syncd
LV Size (Mbytes) 92160
Current LE 2880
Allocated PE 5760
Used PV 3

LV Name /dev/vgprd/lvu14
LV Status available/syncd
LV Size (Mbytes) 15360
Current LE 480
Allocated PE 960
Used PV 2

LV Name /dev/vgprd/lvu151
LV Status available/syncd
LV Size (Mbytes) 2048
Current LE 64
Allocated PE 128
Used PV 2

LV Name /dev/vgprd/lvu152
LV Status available/syncd
LV Size (Mbytes) 2048
Current LE 64
Allocated PE 128
Used PV 2

LV Name /dev/vgprd/lvbkp
LV Status available/syncd
LV Size (Mbytes) 153600
Current LE 4800
Allocated PE 4800
Used PV 2


--- Physical volumes ---
PV Name /dev/dsk/c4t0d0
PV Status available
Total PE 4374
Free PE 0
Autoswitch On

PV Name /dev/dsk/c4t1d0
PV Status available
Total PE 4374
Free PE 0
Autoswitch On

PV Name /dev/dsk/c4t2d0
PV Status available
Total PE 4374
Free PE 610
Autoswitch On

PV Name /dev/dsk/c4t3d0
PV Status available
Total PE 4374
Free PE 150
Autoswitch On

PV Name /dev/dsk/c4t4d0
PV Status available
Total PE 4374
Free PE 950
Autoswitch On


#

[Torsten.]

OK, now we know all 5 external disks are used by the volume group /dev/vgprd.

Do you know how the disks are mirrored?

This is the plan:

- reconfigure 1 HBA(#2) to initiator ID 6
(using mptutil)

- connect each bus with 2 cables to your host (better to shut the host down for this)
e.g.
HBA1 connector1 -> A1 : A2 HBA2 connector1
HBA1 connector2 -> B1 : B1 HBA2 connector2

- now you have 2 device files for each disk

- split the LVM mirror

- remove the device files for the disks to be re-located

- re-locate the disks to the other bus

- create device files

- include the disks to the VG

-include the alternate pathes to the VG

- mirror the lvols


After this, the "vgdisplay -v" would look very similar to this:

--- Physical volumes ---
PV Name /dev/dsk/c2t0d7s2
PV Name /dev/dsk/c4t0d7s2 Alternate Link
PV Status available
Total PE 44940
Free PE 4
Autoswitch On

PV Name /dev/dsk/c8t0d7s2
PV Name /dev/dsk/c6t0d7s2 Alternate Link

PV Status available
Total PE 44940
Free PE 4
Autoswitch On

[Torsten.]

Correction (more clear):

OK, now we know all 5 external disks are used by the volume group /dev/vgprd.

Do you know how the disks are mirrored?

This is the plan:

- reconfigure 1 HBA(#2) to initiator ID 6
(using mptutil)

- connect each bus with 2 cables to your host (better to shut the host down for this)
e.g.
HBA1 connector1 -> A1 : A2 HBA2 connector1
HBA1 connector2 -> B1 : B2 HBA2 connector2

- now you have 2 device files for each disk

- remove the LVM mirror for each LVOL in this VG

- remove the device files for the (empty) disks to be re-located

- re-locate the (empty) disks to the other bus

- create device files

- include the disks to the VG

-include the alternate pathes to the VG

- mirror the lvols


After this, the "vgdisplay -v" would look very similar to this:

--- Physical volumes ---
PV Name /dev/dsk/c2t0d7s2
PV Name /dev/dsk/c4t0d7s2 Alternate Link
PV Status available
Total PE 44940
Free PE 4
Autoswitch On

PV Name /dev/dsk/c8t0d7s2
PV Name /dev/dsk/c6t0d7s2 Alternate Link

PV Status available
Total PE 44940
Free PE 4
Autoswitch On

[Fabian Crea]

Ok, thank you very much, I will work in this.

[Fabian Crea]

HI Torsten, Can I your help, pls ?
I have problem with my proveer and I need do this only I.
Can you send my the sentence and command for to do the change on server, Pls excuseme my english, this is poor.


This is the action plan that you send me.

- reconfigure 1 HBA(#2) to initiator ID 6
(using mptutil)

- connect each bus with 2 cables to your host (better to shut the host down for this)
e.g.
HBA1 connector1 -> A1 : A2 HBA2 connector1
HBA1 connector2 -> B1 : B2 HBA2 connector2

- now you have 2 device files for each disk

- remove the LVM mirror for each LVOL in this VG

- remove the device files for the (empty) disks to be re-located

- re-locate the (empty) disks to the other bus

- create device files

- include the disks to the VG

-include the alternate pathes to the VG

- mirror the lvols

Thanks very much, for all.
Fabian

[Guillermo Rossney]

Hi Torsten!

At last, our provider help us on this, following in some way your suggested plan.

Could you please give us some thoughs about this outputs?

(ioscan -fn and vgdisplay -f)

We can't find the "alternate paths"


Thanks in advance!

[Torsten.]

From your ioscan I can see 2x 5 external disk devices, but without knowledge about the canling nobody can say if there are 5 disks with 2 pathes each or even 10 disks.

You can either tell more about the current cable connections or run a vgscan in preview mode (vgscan -p ... see man vgscan) to get more information.

[Guillermo Rossney]

We have, for sure, only 5 discs in the MSA 30.

vgscan -p output is:

vgscan: Couldn't access the list of physical volumes for volume group "/dev/vg0.
vgscan: The physical volume "/dev/dsk/c2t0d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c2t1d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c3t2d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c7t3d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c7t4d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c4t0d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c4t1d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c4t2d0" is already recorded in the "/etc/.
Couldn't stat physical volume "/dev/dsk/c0t0d0":
Invalid argument




The Volume Group /dev/vg01 was not matched with any Physical Volumes.

[Torsten.]

How about the cabling?
Run vgscan with all parameters:
/usr/sbin/vgscan -a -p -v



The device /dev/dsk/c0t0d0 is the CD-ROM - the message is pretty normal.


The devices
/dev/dsk/c4t0d0
/dev/dsk/c4t1d0
/dev/dsk/c4t2d0

belonging to /dev/vgprd

/dev/dsk/c7t3d0
/dev/dsk/c7t4d0

are in use by /dev/vgbkp

So it looks like
/dev/dsk/c4t3d0
/dev/dsk/c4t4d0
are alternates for /dev/vgbkp

/dev/dsk/c7t0d0
/dev/dsk/c7t1d0
/dev/dsk/c7t2d0
for /dev/vgprd

Check this with the vgscan and vgextend the devices to the appropriate vg.

Check if there are files in
/dev/vg01

[Guillermo Rossney]

As far as i know, the cabling was made following your suggestion

HBA1 connector1 -> A1 : A2 HBA2 connector1
HBA1 connector2 -> B1 : B2 HBA2 connector2


vgscan output is:

vgscan: The physical volume "/dev/dsk/c2t0d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c2t1d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c3t2d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c7t3d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c7t4d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c4t0d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c4t1d0" is already recorded in the "/etc/.
vgscan: The physical volume "/dev/dsk/c4t2d0" is already recorded in the "/etc/.
Couldn't stat physical volume "/dev/dsk/c0t0d0":
Invalid argument


/dev/vg00
/dev/dsk/c2t0d0
/dev/dsk/c2t1d0
/dev/dsk/c3t2d0

/dev/vgprd
/dev/dsk/c7t0d0
/dev/dsk/c7t1d0
/dev/dsk/c7t2d0
/dev/dsk/c4t0d0
/dev/dsk/c4t1d0
/dev/dsk/c4t2d0

/dev/vgbkp
/dev/dsk/c7t3d0
/dev/dsk/c7t4d0
/dev/dsk/c4t3d0
/dev/dsk/c4t4d0

and inside /dev/vg01, only:

crw-r--r-- 1 root sys 64 0x020000 Mar 12 04:33 group

[Torsten. · Accepted solution]

I noticed you have changed the SCSI controller ID as suggested.

It looks good, just vgextend your VGs like mentioned:

vgextend /dev/vgprd /dev/dsk/c4t0d0 /dev/dsk/c4t1d0 /dev/dsk/c4t2d0

vgextend dev/vgbkp /dev/dsk/c7t3d0 /dev/dsk/c7t4d0
and check with "vgdisplay -v"

If not used, delete /dev/vg01
(maybe from a former created and not properly removed volume group)

[Guillermo Rossney]

Torsten, your help was unmeasurable.
Thanks for all.
(Tomorrow Fabian will assign points to your answers)


Just a little doubt:

This sentence will create alternate paths to PV's on /dev/vgbkp

vgextend /dev/vgbkp /dev/dsk/c7t3d0 /dev/dsk/c7t4d0

But to perform the same on /dev/vgprd must not be

vgextend /dev/vgprd /dev/dsk/c7t0d0 /dev/dsk/c7t1d0 /dev/dsk/c7t2d0

instead

vgextend /dev/vgprd /dev/dsk/c4t0d0 /dev/dsk/c4t1d0 /dev/dsk/c4t2d0 ??


Thanks, again.

[Torsten.]

You are right regarding the vgextend command - my fault.
But the command will tell you the disks already included in this VG.

I see in your ioscan you adjusted only 1 HBA to ID 6 - you must adjust the other adapter too.

Already adjusted:
ctl 9 0/3/1/1.6.0 sctl CLAIMED DEVICE Initiator

("6" is the initiator ID)

Make the same for /dev/mpt6 to prevent problems!

[Guillermo Rossney]

Not sure (Fabian was there), but it seems that technician tried to modify de initiatior-id to 6, but the system hangs.

At man pages of mptconfig we see

WARNINGS
Setting the initiator ID on a SCSI controller must be done with caution, since duplicate SCSI IDs on a SCSI bus can result in bus hangs, system hangs, and system panics. The mptconfig command will attempt to verify that no targets are using the specified ID before the change is allowed; however, in the case of multi-initiator (shared bus) configurations, the mptconfig command cannot detect the presence of other initiators on the SCSI bus. Use the ioscan command on any servers connected to the same SCSI bus to verify that the init_id selected is free before using the -i option.


It is safe and necessary to assign the same init-id for both controllers??

[Torsten.]

For 2 HBAs on the same bus one of them MUST be set to ID 6 - otherwise you will get an address conflict (problems like BUS hang etc.).

[Guillermo Rossney]

Thanks Torsten.

Just one of them must be set to 6? (as already we have), or both?

[Guillermo Rossney]

News regarding multipath, from our provider:

"As of HP-UX 11i Version 3, HP recommends using the multipathing inherent in
the mass storage stack, rather than LVM's alternate links. If you want to
use LVM's alternate links, you must use legacy device special files for
physical volumes, and disable the mass storage stack multipathing for those
legacy device special files, using the *scsimgr* command."