cancel
Showing results for 
Search instead for 
Did you mean: 

Drive Path Down

H_Sharma
Level 6

Hello Experts,

We have one Master windows server in cluster and 6 media servers all are in windows.

I could see 1 Drive is in mixed mode and 1 path out of 5 are showing Down TLD which is showing Missing_path: (2,0,15,0) 0007885364 from 1 of the media server.

I tried to up the path on master server by right clicking on the Downed drive but its still showing missing path.

Please help.

11 REPLIES 11

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
You need to troubleshoot on the media server. Missing Path means that the media server OS has lost connectivity to the drive. What does 'scan' command show? Also check for device errors in Event Viewer logs on the media server.

H_Sharma
Level 6

While up the path it shows Unable to up the selected, drive Unknown drive name 297

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
Please read through my post above.

mph999
Level 6
Employee Accredited

As Marianne suggests, the OS has lost connectivity to the drive, you need to fix this - either the drive has disappearsed, or the path has changed.  Delete the drive from NBU and readd, once you have confirmed the OS can see the drive.

I would recommend you enable persistent binding, this will prevent the OS from discovering the drives with different paths - this is not a NBU activity, you need to consult the HBA documentation as it is set via the HBA.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
Repeated drive issues in this environment. Especially with Missing Drives. Previous post with many other similar posts also referenced : https://www.veritas.com/community/forums/drive-issue-3 If I am not mistaken, your clustered master server nodes are the only Windows servers in this environment?

H_Sharma
Level 6
Hi Marianne, Yes its a clustered Master server we had failover 2 days back and still this issue in another node as well? So as in both the cluster node issue is same.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

You have been experiencing issues for a long time now.

How many drives in Windows Device Manager?
What does scan show?

You may want to revisit your previous form posts as the same advice still applies.

***EDIT***

Have you tried to delete all drives and re-run Device Config Wizard on this node?

Extremely important that each of the cluster nodes show separately in Device database as they will never have exactly same device paths at OS-level.
The Storage Unit needs to be configured with virtual hostname.

Have you enabled SCSI-3 reservation and Persistent Binding as per advice in previous posts?

H_Sharma
Level 6

Hello Marianne,

We logged a case with Veritas and TSE investigated OS has lost connectivity to the Drive Once OS can see the drive run tpautoconf -a ?

My queries are as below.

1:- We have 40 drives how should i know if one of the specific drive has lost connctivity to OS?

2:- tpautoconf -a will this command restart netbackup services on master server as we have master in cluster?

    The server question is a media server and also Robot control Host?

Please help

mph999
Level 6
Employee Accredited

1.  You will see missing path in tpconfig -d

2.  No, I'd suggest not runnig that command as it reconfigures everything that the server is connected to where the command is run.

Miossing path is not a NBU issue, it is a SAN/ OS issue and you need to resolve the issue at the SAN / OS level, then NBU will have no problem.  Look into persistent binding which is set via the HBA.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

1. Missing Path to any kind of device will be logged in Windows Event Viewer System log. You can talk to Microsoft about Event Viewer monitoring and reporting tools.
Maybe your question should be 'what is causing this' and 'how to prevent this'.
You need to look at device drivers (tape drive and hba) and driver settings. 
Some driver settings don't 'play nice' in a shared environment and will lose connectivity when another server has reserved the drive.
Log calls will tape drive vendor as well as hba vendor for their recommendations.

2. Freeze the Cluster or Cluster Group before reconfiguring devices to prevent a failover when services are restarted.

PS: 

Any reason why you keep on ignoring this advice (in this and previous discussions)?:

Have you enabled SCSI-3 reservation and Persistent Binding as per advice in previous posts?

H_Sharma
Level 6

Hello Experts Issue is resolved by reconfiguing the device. I can not close this post. Please mark this as solution.