serviceguard setup error on cmapplyconf

[Amir Haboosheh]

i am trying to setup a 2 node cluster. when i run:
# cmapplyconf -k -v -C /etc/cmcluster/clust1.config

i get this:
----------------BEGIN-------------------
Begin cluster verification...
Checking cluster file: /etc/cmcluster/clust1.config
Note : a NODE_TIMEOUT value of 2000000 was found in line 130. This value
is recommended if the top priority is to reform the cluster as fast
as possible in case of failure. If the top priority is to minimize
reformations, consider using a higher setting. For more information see
the cluster configuration ASCII file or the Managing Serviceguard manual.
Checking nodes ... Done
Checking existing configuration ... Done
Gathering storage information
Found 1 devices on node unix1
Found 1 devices on node unix2
Analysis of 2 devices should take approximately 1 seconds
0%----10%----20%----30%----40%----50%----60%----70%----80%----90%----100%
Found 1 volume groups on node unix1
Found 1 volume groups on node unix2
Analysis of 2 volume groups should take approximately 1 seconds
0%----10%----20%----30%----40%----50%
Timed out trying to query the following volume group(s) on node unix1
/dev/vg01
----60%----70%----80%----90%----100%
Gathering network information
Beginning network probing (this may take a while)
Completed network probing
Checking for inconsistencies
Maximum configured packages parameter is 150.
Configuring 0 package(s).
150 package(s) can be added to this cluster.
200 access policies can be added to this cluster.
Creating the cluster configuration for cluster cluster1
Adding node unix1 to cluster cluster1
Adding node unix2 to cluster cluster1
WARNING: Cluster lock on disk /dev/dsk/c2t0d1 is missing! Until it is fixed, a single failure could cause all nodes in the cluster to crash.
/dev/dsk/c2t0d1 is inaccessible
error

Failed to check cluster lock /dev/vg01:/dev/dsk/c2t0d1
The volume group /dev/vg01 needs to be activated in order to initialize the cluster lock on it.
cmapplyconf: Unable to apply the configuration
------------------- END -----------------

vg01 is also my cluster lock.
when i run cmcheckconf everything checks ok and i don't get any errors.

thanks,
amir

[Kenan Erdey]

Hi,

first activate vg01 (vgchange -a y /dev/vg01) then try cmapplyconf. After it works deactivate it with -a n and start the cluster.

Kenan.

[Amir Haboosheh]

i did that already before running the cmapplyconf. but thanks.

[Kenan Erdey]

Hi,

firstly we can check disk is ok with:

vgdisplay -v vg01
diskinfo /dev/rdsk/c2t0d1
pvdisplay /dev/dsk/c2t0d1
dd if=/dev/rdsk/c2t0d1 of=/dev/null bs=2048k

[Asif Sharif]

Salam Amir,

The message below may indicate that disk is not accessible.

WARNING: Cluster lock on disk /dev/dsk/c2t0d1 is missing!

ioscan -fnC disk--> see any unclaimed device

vgdisplay -v vg01
diskinfo /dev/rdsk/c2t0d1

Regards,
Asif Sharif

[Sharma Sanjeev]

Hi Amir

First you have to check the Health of disk
/dev/dsk/c2t0d1 and vg01 belongs to this disk.

Regards
Sanjeev

[Amir Haboosheh]

hey gues, tanks for your replies. i have an open case with hp on this and they determined (after many days and logs and tests) that the problem is at the Storage or LVM layer.

my setup:
2x rx2660 FC conected to an MSA1500 with dual controllers. the MSA1500 has 1x MSA30 connect to it. both servers are connected to the MSA1500 with FC.

is this setup correct or do i need another MSA1500 (one for each server) and each MSA1500 will then be connected to the MSA30?

any help is appriciated, thanks you.

amir

[Pedro Cirne]

Hi,

Would post here the output of:

#ioscan -fnCdisk
#pvdisplay /dev/dsk/c2t0d1

Pedro

[Deepak Kr]

Hi Amir,

I agree with all comments given here. As a first plan of action you please verify if disk is visible on both of your systems and also you can mount it before you actually go for running cmapplyconf next time.

Thank!

[Amir Haboosheh]

ok,
thank you guys for all the replies.
after 2 weeks of back and forth with hp serviceguard/LVM/Storage departments i found out that an MSA30 connected to an MSA1500 (w/ dual MSA1000 controllers) connected to an integrity server running hp-ux must be in an active/active firmware v.7 on the MSA1000 for it to work. active/passive does is not compatible unless you have a fiber switch in between.

thanks guys,
amir

[Torsten.]

Not exactly.

http://h18000.www1.hp.com/products/quickspecs/11945_na/11945_na.HTML

"OS Support for
Single controller or dual controller configurations with active/passive firmware:

HP-UX 11.11 and 11.23 (single controller in homogeneous environments only)

...

HP-UX support
(specific guidelines)

The MSA1500 for HP-UX supports dual active/active controller or single controller configurations (Active/passive configurations are not supported). HP-UX single controller support is enabled with the standard firmware that ships with the MSA1500HP-UX implemented with dual active controllers requires use of firmware available through a free download off the MSA1500 web site.
NOTE: all MSA1500 HP-UX starter kits ship with active/active firmware pre-installed, which supports both single controller or active/active controller configurations.


ACU/CLI (Command Line Interface) is used for configuration and Event Monitoring Service (EMS) for monitoring.
Root/Boot/Dump/Swap/Install is also supported.
PVlinks is supported for multi (HBA) path / single controller (2 HBAs) and dual/active controller configurations.
Fabric and direct connect are supported
HP-UX 11.11 & 11.23, PA-RISC and Itanium "