cancel
Showing results for 
Search instead for 
Did you mean: 

vmoprcmd show down drive but not vmoprcmd -d

Zilf
Level 3
Partner Accredited

hello all

here is the result of a vmoprcmd (no option)

adic47_F002BA200D No No No hcart
alpha /dev/nst0 TLD
beta /dev/nst2 DOWN-TLD

adic47_F002BA2010 Yes Yes 606872 606872 Yes hcart-Clean
alpha /dev/nst5 TLD
beta /dev/nst7 ACTIVE

adic47_F002BA2013 Yes Yes 606861 606861 Yes hcart
alpha /dev/nst1 ACTIVE
beta /dev/nst3 DOWN-TLD

2 drives on media server beta are down. However, by doing the same command with option -d :

Drv Type Control User Label RecMID ExtMID Ready Wr.Enbl. ReqId
0 hcart TLD Yes 606875 606875 Yes Yes 0
1 hcart TLD Yes 606872 606872 Yes Yes 0
2 hcart TLD - No - 0
3 hcart TLD - No - 0
6 hcart TLD Yes 604554 604554 Yes Yes 0
7 hcart TLD Yes 604549 604549 Yes Yes 0
8 hcart TLD - No - 0
9 hcart TLD Yes 604551 604551 Yes Yes 0

everything is up. Same thing with a tpconfig -d on the beta media server : 

Id DriveName Type Residence
Drive Path Status
****************************************************************************

2 adic47_F002BA2013 hcart TLD(47) DRIVE=2
/dev/nst3 UP
3 adic47_F002BA200D hcart TLD(47) DRIVE=1
/dev/nst2 UP

so what happens doc ?

7 REPLIES 7

Marianne
Level 6
Partner    VIP    Accredited Certified

 

 

I would restart ltid on beta to get everything in sync again.

There could be intermittent comms issue with the master so that status is not updated on the master.

Zilf
Level 3
Partner Accredited

no didn't work. The two drives are still be shown as down.

what's the difference between vmoprcmd and vmoprcmd -d ? one will use the globdb and the other will ask directly to the media server ?

I just DOWNed a drive on a media server and it shows correctly in vmoprcmd with or without the -d

Make sure you specify the media server  vmoprcmd -d -h beta

 

Marianne
Level 6
Partner    VIP    Accredited Certified

Suspicion is then with the master server... not polling media servers often enough or something to that effect. 

As far as I know vmoprcmd info is held in memory and not in config files or database... 
There used to be a TN about this (where vmoprcmd output was different from tpconfig) but it does not seem to exist anymore.

If this is a real problem to you, try to restart NBU on the master. 

 

Zilf
Level 3
Partner Accredited

well, it is far easier to manage masters from one server only and ssh the vmoprcmd command than trying to find the media servers that are up and doing the tpconfig -d (or vmoprcmd -d -h).

As is, I will agree that restarting the master is a workaround but unfortunately, I cannot restart the master anytime I want.

What would be interesting is to know how we can refresh this vmoprcmd command.

Marianne
Level 6
Partner    VIP    Accredited Certified

My experience has been the same as @Will_Restore - vmoprcmd works correctly on the master server.

So, IMHO, the issue is not with NBU but with something in your environment causing the master to not refresh device status on media servers... 

You may want to log a Support call for assistance with troubleshooting.

Hello,


The 'vmoprcmd' command will give the output of devices connect all media servers within that environment.

Also it will show the status of media server e.g active-disk, active etc.


The 'vmoprcmd -d' command will give the devices output of that local media server.

I think the vmoprcmd command connect to vmd service to get the information of devices.


For quick check we can try to stop/start the vmd service using below command


To stop:
/usr/openv/volmgr/bin/vmd -stop / kill -9 <vmd pid>


To start:
/usr/openv/volmgr/bin/vmd -v

Note: Stop/Start vmd service will affect running jobs on that media server.


Enable the daemon logs on problematic media server to troubleshoot further.

Thanks & Regards,