Re: SAN Drive Presentation


"GMKovach" <GMKovach@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:2C3C58E6-A68F-40FA-A4A8-AFB072966AE6@xxxxxxxxxxxxxxxx
I mistakenly was comparing "paths" instead of drives. On the File&Print
cluster they are in the right order as far as MPIO is concerned, but not
the
same in Disk Manager. This is how they compare:

Disk# Node1 Node2
0 C:\ C:\
1 Quorum Q: DHCP_Data D:
2 File_Data F: Print_Spooler P:
3 Unknown Unknown
4 Unknown Unknown
5 vbe_data V: Unknown

this is normal, 3 disks are online on Node 1... which are "Unknown" on Node
2... (and visa versa for the other 2 Disks)
this is normal behaviour on a cluster


I discovered this issue when I go to run a backup using Backup Exec and I
get all these VSS errors on the drive:
Event ID 12290
Volume Shadow Copy Service warning:

GetVolumeInformationW(\\?\Volume{6335ed8b-2ecf-11da-9bb4-0014384d9365}\,NULL
,0,NULL,NULL,[0x00000000], , 260) == 0x00000057. hr = 0x00000000.


I cannot comment on this, as I don't know enough about VSS to go into detail
on this.

When I try to fail it over I get these drive corruption errors:

Event Type: Error
Event Source: ClusSvc
Event Category: Physical Disk Resource
Event ID: 1066
Date: 10/26/2006
Time: 4:48:20 PM
User: N/A
Computer: FP1
Description:
Cluster disk resource "DHCP Physical Storage" is corrupt. Run 'ChkDsk /F'
to
repair problems. The volume name for this resource is
"\\?\Volume{6335ed90-2ecf-11da-9bb4-0014384d9365}\".
If available, ChkDsk output will be in the file
"C:\WINDOWS\Cluster\ChkDsk_Disk4_Sig037C3AE1.log".
ChkDsk may write information to the Application Event Log with Event ID
26180.

For more information, see Help and Support Center at
http://go.microsoft.com/fwlink/events.asp.

Event Type: Error
Event Source: ClusSvc
Event Category: Physical Disk Resource
Event ID: 1035
Date: 10/26/2006
Time: 4:48:20 PM
User: N/A
Computer: FP1
Description:
Cluster disk resource 'DHCP Physical Storage' could not be mounted.

For more information, see Help and Support Center at
http://go.microsoft.com/fwlink/events.asp.

Event Type: Error
Event Source: ClusSvc
Event Category: Failover Mgr
Event ID: 1069
Date: 10/26/2006
Time: 4:48:20 PM
User: N/A
Computer: FP1
Description:
Cluster resource 'DHCP Physical Storage' in Resource Group 'DHCP Server'
failed.

For more information, see Help and Support Center at
http://go.microsoft.com/fwlink/events.asp.


Are you sure this "DHCP Physical Storage" Disk is presented to the second
node correctly (read/write, all access OK)

reason I ask; is the sequence of events, within the same second you get a
"Event ID: 1066" which should start chkdsk (and devices go online-pending)
but within the same second you get your 1035 and 1069, which could mean that
the disk is not even readonly.. or not ready or any other state your on your
SAN.... "as if the disk is not accessible, not even by chkdsk". I would have
expected that a chkdsk take a little more time than 1 second

Are you sure you could access this disk from the second node without cluster
installed ?


I just tried to fail over another group (Print Spooler) and I get the same
errors. What are the chances of corrupting 2 seperate drives at the same
time?

I'm thinking that when I try to failover Drive1 on Node2 it sees that
Drive1
on Node1 is already occupied, hence the failure. just a thought. any ideas
on
a fix?

cluster will bring online through disk-signature, not physical disk
numbering


Drive #5 (V:) will fail over normally (Veritas Group).

Coincidently, on my Exchange cluster, Disk Manager mirrors betwee the two
nodes; meaning what's active on node1 is inactive on node2.

I guess on that cluster your physical disk numbering is the same between
nodes, but this should not have any affect on your cluster, cluster will
bring disks online based on signature, not on physical disk numbering.


Thanks

GMKovach


"Edwin vMierlo" wrote:

Physical drive numbering does not have to be the same between nodes
(it would be nice for reference, but not needed)

is it all physical disks or just some ?

if you completely power-down the physical node 1, does anything failover
to
node 2 ?
(unless this is production, then obviously do not do a powerdown)

Next step is to look in the system event-log for any messages, or the
cluster log for any clues


"GMKovach" <GMKovach@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:E5283943-5EBD-4700-9D28-F35339AA3875@xxxxxxxxxxxxxxxx
Hi,
I'm currently having an issue with physical disks not failing over
in
an
Active/Passive 2003 Cluster. The drives are on an HP EVA6000 SAN. When
I
run
the MPIO utility I notice that the nodes do not see the devices
(Disks) in
the same order. Is this an issue? Also, if you compare Disk Manager
between
the nodes, should they be opposite of each other? For example, if
Disk1 on
Node1 is Active, Should Disk1 on Node2 be inactive?

Thanks

GMKovach





.

Relevant Pages

  • SUMMARY: changed WWID on cluster member boot disk
    ... disk and quorum disk of a single-member cluster, ... I could no longer boot from the cluster disks, ... the pre-cluster stand-alone system disk; ... the root1_domain on LUN containing the member boot disk was found ...
    (Tru64-UNIX-Managers)
  • Re: Single node cluster physical disk issue
    ... single node cluster?? ... I am experiancing an issue with "Bring Online" a physical disk resource. ...
    (microsoft.public.windows.server.clustering)
  • Re: Creating a wide area VMS Cluster
    ... > My goal is to provide a disaster tolerant cluster for both OS and data. ... disrupting the balance of the effect of votes between sites A and B. ... You have the option of a single shadowed system disk between the ...
    (comp.os.vms)
  • RE: Cluster IP Address Does not fail over
    ... The cluster IP has no dependicies at all. ... Node1 disk manager sees LUN5. ... [DiskArb] ...
    (microsoft.public.windows.server.clustering)
  • Join an existing cluster
    ... I had a cluster setup with 2 computers running windows ... shared disk array. ... Creating a dummy Local Quorum resource. ... on the same storage bus as the boot disk... ...
    (microsoft.public.windows.server.clustering)