cancel
Showing results for 
Search instead for 
Did you mean: 

SFCFSHA 6.2 : DG in unknown status

stephane4tests
Level 2

Hi all, On a 2 nodes system I did a fresh installation of SFCFSHA 6.2.
There's an IBM array with shared Luns. There's only one path.

Linux are 6.7 64bit built identically.
After installation, llt failed to start.

So I installed also the MR and patch : sfha-rhel6_x86_64-MR-6.2.1 sfha-rhel6.7_x86_64-Patch-6.2.1.100
Then llt is starting, the cluster is up and I finished to configure it with disk fencing.
Good.
But as soon as I add a DG and the FS (only one on this dg) resources in VCS, the DG appears with a question mark and the "unknown status".
And I see this error message coming repeatedly :
2016/02/18 15:29:22 VCS ERROR V-16-2-13040 (mynode1) Resource(cvmvoldg1): Program(/opt/VRTSvcs/bin/CVMVolDg/monitor) was abnormally terminated with the exit code(0x8b).

Following a post about this error I verified the files in /etc/VRTSvcs/conf/types.cf and /etc/VRTSvcs/conf/config/types.cf and copied the most recent one on the other (stopping cluster, copying files, starting cluster).
No positive results.

Trying to start the resource group lead to a message saying the group cannot be starting since it had probes pending on both nodes. I also tried with one node with no more success.

Any advices ?

Here's some informations :
# vxdg list
NAME STATE ID dg_data1 enabled,shared,cds 1455640017.22.mynode1
# cfsdgadm display
Node Name : mynode2 DISK GROUP ACTIVATION MODE dg_data1 sw Node Name : mynode1 DISK GROUP ACTIVATION MODE dg_data1 sw
# cfsmntadm display
Cluster Configuration for Node: mynode2 MOUNT POINT TYPE SHARED VOLUME DISK GROUP STATUS /database/dg_data1 Regular v_dg_data1 dg_data1 NOT MOUNTED Cluster Configuration for Node: mynode1 MOUNT POINT TYPE SHARED VOLUME DISK GROUP STATUS /database/dg_data1 Regular v_dg_data1 dg_data1 NOT MOUNTED

Thanks

1 ACCEPTED SOLUTION

Accepted Solutions

stephane4tests
Level 2

Hi Doug,

I  even created a new DG ! But no matter : at last I did something it worked !

Following are the steps that solved this issue. Actually I don't know if the first step is necessary but it's not a bad idea to do it.

1/ upgrade  RHEL 6.7 with latest patches from  RHEL (yum update).

2/ Install Veritas and its patches in the following order without configuring the cluster and so the processes are NOT started :

VRTS_SF_HA_Solutions_6.2_RHEL

glm-rhel6_x86_64-Patch-6.2.0.100            Patch                                   6.2.0.100             6.2                2015-08-24

sfha-rhel6_x86_64-MR-6.2.1                       Maintenance Release      6.2.1                     6.2                2015-04-24

fs-rhel6_x86_64-Patch-6.2.1.100               Patch                                   6.2.1.100             6.2                2015-09-02

sfha-rhel6.7_x86_64-Patch-6.2.1.100       Patch                                   6.2.1.100             6.2                2015-09-14

vcscvm-rhel6_x86_64-Patch-6.2.1.100     Patch                                   6.2.1.100             6.2                2015-09-15

cpi-linux-Patch-6.2.1.700                             Patch                                   6.2.1.700             6.2.1                2015-12-29

Now configure the cluster

               /opt/VRTS/install/installsfrac62  -configure

(and the fencing if any with  –fencing )

That's  all.

View solution in original post

4 REPLIES 4

Doug_Snyder
Level 3
Partner

Hi Stephanie,

Have you tried deporting the DB and re-importing it as a non-shared DG?  If so, does the volume start then?

stephane4tests
Level 2

Hi Doug,

I  even created a new DG ! But no matter : at last I did something it worked !

Following are the steps that solved this issue. Actually I don't know if the first step is necessary but it's not a bad idea to do it.

1/ upgrade  RHEL 6.7 with latest patches from  RHEL (yum update).

2/ Install Veritas and its patches in the following order without configuring the cluster and so the processes are NOT started :

VRTS_SF_HA_Solutions_6.2_RHEL

glm-rhel6_x86_64-Patch-6.2.0.100            Patch                                   6.2.0.100             6.2                2015-08-24

sfha-rhel6_x86_64-MR-6.2.1                       Maintenance Release      6.2.1                     6.2                2015-04-24

fs-rhel6_x86_64-Patch-6.2.1.100               Patch                                   6.2.1.100             6.2                2015-09-02

sfha-rhel6.7_x86_64-Patch-6.2.1.100       Patch                                   6.2.1.100             6.2                2015-09-14

vcscvm-rhel6_x86_64-Patch-6.2.1.100     Patch                                   6.2.1.100             6.2                2015-09-15

cpi-linux-Patch-6.2.1.700                             Patch                                   6.2.1.700             6.2.1                2015-12-29

Now configure the cluster

               /opt/VRTS/install/installsfrac62  -configure

(and the fencing if any with  –fencing )

That's  all.

View solution in original post

Doug_Snyder
Level 3
Partner

Glad to hear it is resolved.  Sometimes the patching can trip you up (it is much better in 7.0).  I reccomend using an assessment https://sort.veritas.com/data_collectors prior to any install or upgrade.  The assessment tool runs in just a few minutes and will let you know if there are any required Veritas or OS patches.  Take care.

paiih
Level 1
Partner

I have the same problem.

Finally, find the solution:
Support for Storage Foundation High Availability (SFHA) Solutions and InfoScale on Red Hat Enterprise Linux 6.7 (RHEL 6.7)
http://www.veritas.com/docs/000025669

workaroud: replace an old version ksh