Extendvg problem

[ereshatezer]

Hi @all,
I read through this forum, but unfortunately there was no solution for my problem. Here it is:


--> extendvg K02sapvg hdisk9
0516-024 extendvg: Unable to open physical volume.
Either PV was not configured or could not be opened. Run
diagnostics.
0516-792 extendvg: Unable to extend volume group.
0516-1254 extendvg: Changing the PVID in the ODM.
Method error (/usr/lib/methods/chgdisk):
0514-047 Cannot access a device.

0516-024 extendvg: Unable to open physical volume.
Either PV was not configured or could not be opened. Run
diagnostics.
0516-792 extendvg: Unable to extend volume group.
0516-792 extendvg: Unable to extend volume group.
***
And some additional info:
***
hdisk0 005d07caf025308d rootvg
hdisk1 005d07cac7e3c5fd rootvg
hdisk2 005d07ca4eed0651 VG
hdisk3 none None
hdisk4 005d07ca684c1848 VG
hdisk5 none None
hdisk6 005d07ca39edcd2f vg
hdisk7 005d07cacb3b6628 VG
hdisk8 005d07ca4dbe4c1c VG
hdisk9 none None
hdisk10 none None
hdisk11 005d07ca4eed07ac VG
hdisk12 none None
hdisk13 005d07ca68ce550f VG
hdisk14 none None
hdisk15 none None
hdisk16 none None
hdisk17 005d07ca4dbe4d94 VG
***
--> errpt -a
---------------------------------------------------------------------------
LABEL: SC_DISK_ERR2
IDENTIFIER: 79B0DF89

Date/Time: Thu Feb 16 12:20:31 NFT
Sequence Number: 254969
Machine Id: 005D07CA4C00
Node Id: rshama23
Class: H
Type: PERM
Resource Name: hdisk9
Resource Class: disk
Resource Type: osdisk
Location: 11-08-01
VPD:
Manufacturer................LSI
Machine Type and Model......INF-01-00
ROS Level and ID............30353330
Device Specific.(Z0)........000003321F004032

Description
DISK OPERATION ERROR

Probable Causes
DASD DEVICE

Failure Causes
DISK DRIVE
DISK DRIVE ELECTRONICS

Recommended Actions
PERFORM PROBLEM DETERMINATION PROCEDURES

Detail Data
SENSE DATA
0A00 2800 0000 0000 0000 0800 0000 0000 0000 0000 0000 0000 0102 0000 7000 0500
0000 0098 0000 0000 9401 0000 0000 0000 0100 0000 0000 0000 0000 0000 0000 880B
0099 9999 9928 0000 0000 0000 0008 0008 0000 3154 3234 3830 3738 3635 2020 2020
2020 0530 2100 0007 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0005 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 98CE E235 3032 3136 3036 2F30 3530 3835 3000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 2200 0000 0004 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000
***
-rmdev + cfgmgr does not work.
-none of the disks usable.

Any ideas?
Thx in advance!

 

[mrn]

Is the VG mirrored? If so you need to add two disks in not you could try formatting the disk.

Have you tried to create a new vg on the disk?

Mike

"A foolproof method for sculpting an elephant: first, get a huge block of marble, then you chip away everything that doesn't look like an elephant."

http://www.airport-parking-site.co.uk/

 

[ereshatezer]

mkvg is not possible and the VG is not mirrored.
*****
0516-024 mkvg: Unable to open physical volume.
Either PV was not configured or could not be opened. Run
diagnostics.
0516-862 mkvg: Unable to create volume group.
0516-1254 mkvg: Changing the PVID in the ODM.
Method error (/usr/lib/methods/chgdisk):
0514-047 Cannot access a device.

0516-024 mkvg: Unable to open physical volume.
Either PV was not configured or could not be opened. Run
diagnostics.
0516-862 mkvg: Unable to create volume group.
0516-304 mkvg: Unable to find device id in the Device
Configuration Database.
0516-324 mkvg: Unable to find PV identifier for physical volume.
The Device Configuration Database is inconsistent for physical
volumes.

 

[mrn]

Try a format on the disk then.

It's under

Smit diag
Current Shell
Task Selection
Format media

Mike

"A foolproof method for sculpting an elephant: first, get a huge block of marble, then you chip away everything that doesn't look like an elephant."

http://www.airport-parking-site.co.uk/

 

[p5wizard]

Before you start formatting, I'd find out if those seemingly 'free' disks aren't in use by another server.

What are those disks anyway?

lscfg output?

HTH,

p5wizard

 

[ereshatezer]

--> lscfg -vpl hdisk14
DEVICE LOCATION DESCRIPTION

hdisk14 21-08-01 Other FC SCSI Disk Drive

Manufacturer................LSI
Machine Type and Model......INF-01-00
ROS Level and ID............30353330
Device Specific.(Z0)........000003321F004032


PLATFORM SPECIFIC

Name: disk
Node: disk
Device Type: block

****
Unfortunately I administer the server remotely, and to be honest, I do not know anything about them.

 

[p5wizard]

Well, someone out there must be in charge of their storage box. Have them verify what's what with regard to LUNs presented to the AIX server.

I have a hunch that you have 2 hdisk names (2 paths) for each LUN being presented (8 hdisks in VG, 8 "free"...)

I'd check with lscfg if there are pairs with the same "Device Specific.(Z0)" value. Though that is not necessarily the case for different paths to the same LUN. Depends on the architexture of the storage box.

What multipathing software (if any) is in use for these LUNs anyway?


HTH,

p5wizard

 

[ereshatezer]

-->for i in `lspv|awk '{print$1}'`
do
echo $i
lscfg -vpl $i|grep Z0
done

hdisk0
Device Specific.(Z0)........000003029F00013A
hdisk1
Device Specific.(Z0)........000003029F00013A
hdisk2
Device Specific.(Z0)........000003321F004032
hdisk3
Device Specific.(Z0)........000003321F004032
hdisk4
Device Specific.(Z0)........000003321F004032
hdisk5
Device Specific.(Z0)........000003321F004032
hdisk6
Device Specific.(Z0)........000003321F004032
hdisk7
Device Specific.(Z0)........000003321F004032
hdisk8
Device Specific.(Z0)........000003321F004032
hdisk9
Device Specific.(Z0)........000003321F004032
hdisk10
Device Specific.(Z0)........000003321F004032
hdisk11
Device Specific.(Z0)........000003321F004032
hdisk12
Device Specific.(Z0)........000003321F004032
hdisk13
Device Specific.(Z0)........000003321F004032
hdisk14
Device Specific.(Z0)........000003321F004032
hdisk15
Device Specific.(Z0)........000003321F004032
hdisk16
Device Specific.(Z0)........000003321F004032
hdisk17
Device Specific.(Z0)........000003321F004032

******
--> odmget -q name=hdisk14 CuDv|grep parent
parent = "fscsi1"
--> odmget -q parent=fscsi1 CuDv|grep name
name = "rmt4"
name = "rmt5"
name = "rmt6"
name = "hdisk10"
name = "hdisk11"
name = "hdisk12"
name = "hdisk13"
name = "hdisk14"
name = "hdisk15"
name = "hdisk16"
name = "hdisk17"

--> odmget -q parent=fscsi0 CuDv|grep name
name = "rmt1"
name = "rmt2"
name = "rmt3"
name = "hdisk2"
name = "hdisk3"
name = "hdisk4"
name = "hdisk5"
name = "hdisk6"
name = "hdisk7"
name = "hdisk8"
name = "hdisk9"
name = "ovpass0"



I asked the storage responsible about the problem. Waiting for answer....

 

[p5wizard]

The "Device Specific.(Z0)" values are all the same for the FC disks - probably meaning that they are all in the same Storage server box. But doesn't give any other info to me about multipathing.

But the fact that you see 3 tapedrives and 8 disks thru each FC adapter only makes that hunch I have a bit stronger...

Better await that answer - and hold off on the formatting!


HTH,

p5wizard

 

[RMGBELGIUM]

execute the following command :

lqueryvg -p hdisk14 -L

Then you will be able to see which LV's reside in the VG of which the disk could be a member...

rgds,

R.

 

[ereshatezer]

--> lqueryvg -p hdisk14 -L
0516-062 lqueryvg: Unable to read or write logical volume manager
record. PV may be permanently corrupted. Run diagnostics


****
I asked the storage responsible who told me: everything is OK with the configuration....

 

[p5wizard]

Well, did (s)he tell you which volumes were assigned to your server? Number of volumes? Size of the assigned volumes? Number of paths possible between your server and the storage server?

HTH,

p5wizard

 

[mrn]

p5 are you suggesting that the disks may be duplicates of the same disk?

I have seen this but the devices show up as pdisk's not hdisk.

Can you provide the output of prtconf

If this server is attached to a SAN can the SAN guys not tell you what LUN's & Zones are assigned to this server?

Mike

"A foolproof method for sculpting an elephant: first, get a huge block of marble, then you chip away everything that doesn't look like an elephant."

http://www.airport-parking-site.co.uk/

 

[mrn]

Also

lsdev -H

Mike

"A foolproof method for sculpting an elephant: first, get a huge block of marble, then you chip away everything that doesn't look like an elephant."

http://www.airport-parking-site.co.uk/

 

[ereshatezer]

Thank you for everyone for the help. At least one of the guy who installed the system could explain me this situation.

He said, quote:

"for every LUN in the disksystem there are two disks seen by the operating system. This is because the disksystem has two internal controllers and just one handles a given LUN. The other controller shows a "phantom disk", which is the same LUN but cannot be accessed directly. So there are no spare disks assigned to the machine.

 

[p5wizard]

That's what I was telling you all along...

A good point to find out now is: are all your "active" disks on the same fcsX/fscsiX adapter or is the IO load balanced over the 2 adapters you have. There may also be a path failover/load sharing driver available.

Based on the lsvg, lscfg and odmget output you posted earlier, at least your IO load looks kind of balanced: (5 active disks on fscsi0 and 3 active on fscsi1)
fscsi0/hdisk2 005d07ca4eed0651 VG
fscsi0/hdisk3 "phantom" None
fscsi0/hdisk4 005d07ca684c1848 VG
fscsi0/hdisk5 "phantom" None
fscsi0/hdisk6 005d07ca39edcd2f VG
fscsi0/hdisk7 005d07cacb3b6628 VG
fscsi0/hdisk8 005d07ca4dbe4c1c VG
fscsi0/hdisk9 "phantom" None
fscsi1/hdisk10 "phantom" None
fscsi1/hdisk11 005d07ca4eed07ac VG
fscsi1/hdisk12 "phantom" None
fscsi1/hdisk13 005d07ca68ce550f VG
fscsi1/hdisk14 "phantom" None
fscsi1/hdisk15 "phantom" None
fscsi1/hdisk16 "phantom" None
fscsi1/hdisk17 005d07ca4dbe4d94 VG

In any case, if you need more space, the SAN guys/dolls at your company are the people to talk to first...

HTH,

p5wizard