Networker : Error reading volume label :read open error, no tape in the specified device

Looks like a well-known issue - on a tape library with multiple drives you are talking to the wrong one.

Most likely because the OS changed the assignment of the tape drives after a reboot/restart. Yes, this was possible. I really have to dig a bit deeper into my archive how to correct that but for the moment I think it is more important to recover the data.

How to find out/address the correct device? - Load the tape into a drive and then run

  scanner -n -m -v device_name

to verify where the tape is located. Do this for the other drives. Then correct your jukebox setup with jbedit .

Another very easy solution would be to temporarily disconnect all tape drives but the first one. Reboot the server and run jbconfig.

A

Hello.. Thanks for your input.

I tried running the scanner command and I have attached the screenshots for some of the things.

Also, I observed that the inquire command outputs only 1 tape drive i.e. Tape0. The 2nd drive is not visible here.
I also checked the same in Server's Device Manager. There also, only 1 Tape drive is showing.

I don't know what is the issue here.

Can you please guide.

The secret is here - as you stated in your opening message:  Media critical event : waiting for LTO Ultrium-5 tape A00010 on server2

So you share the library between your NW server ??? and the remote storage node server2, which unfortunately is a NDMP filer. That's why this tape device is defined as NDMP device.

This means that the tape drive is controlled by the NDMP tape services running on server2. NW still controls the jukebox and this seems to run properly, but it does not control the data sent/read to/from the device. Consequently, you cannot use 'normal' NetWorker commands like 'scanner'.

But from the message it looks like you do not have server2 in service any longer. So NW has successfully done his part - it has loaded the tape into the correct drive. But if the server is gone, he will not be able to mount it for recoveries. He sends a tape mount request but obviously gets no answer any more. (NDMP is very rudimentary. There is no NW software installed to receive better info.) - Am I correct about this image?

How to correct the problem? - In general, there is no other way than setting/starting up the NDMP filer once again because, as I said, only such driver is able handle NDMP data.

There might be another solution:

There was a very old product called SnapImage which - AFAIR - had a separate NDMP tape driver which you may install and use to make a standard Windows host NDMP compatible. It could be worth to try. You better ask Dell/EMC support to provide it.

   Good luck.

Thank you so much for such an elaborated reply.

It seems you are right and I checked in Mgmt Console > Confiugration > Local Hosts

The local host "scl-bkup-svr" is reachanble but "server2" is showing error : 

Unable to connect to Server : Failed to connect using UDP Ping.

But actually I am able to ping server2 using normal ping commands from networker server and the server2 is actually ON and reachable.

But I am not able to find out why it is showing the above error. Any suggestions here..??

Thank you.

'ping' is just a TCP command which tests a network connection. It does not use a specific data transfer protocol. Consequently, any TCP/IP device will be able to answer if there is a connection.

NDMP is a specific transfer protocol, based on a special NDMP OS. Such is designed in a way that one can install any other software like a NW client. Please see the internet for more details.

Furthermore, for a NW client, you must enable your firewall for both ports 7937 & 7938 for both, TCP & UDP protocols. This is obviously not the case.

Can you run a backup of server2 at all? - I do not think so.

BTW - what is the brand/model of your NDMP filer?

Our NDMP filer is a Dell EMC VNX5500 Unified storage.

Ans yes, we are unable to run backups as well as recoveries.

I would now check up the NDMP services and firewall settings on our NDMP filer and update here.

In fact, there is an alternative - the NW Data Service Agent (DSA). He has been available for at least 10 years. The idea is to 'wrap' the NDMP data stream in a way, that it can be sent to a standard (non-NDMP) device. The other benefit is that these media can be scanned as usual.

Unfortunately, this does not help for your existing backups.

May I suggest that you ask support or a VNX User Group how to check/control the NDMP drivers on that storage. This is the prerequisite which is required for a NW NDMP configuration.

Hello, I have not used NetWorker in this way but I have used VNX as an NDMP source with a shared FC tape library.

Have you checked the VNX is able to see the tape library and drive?

server_devconfig server_2 -list -scsi -nondisks -all

if you cannot see any mention of library/drive then you may have to run a scan to tell VNX to discover it:

server_devconfig server_2 -create-scsi -nondisks -all

Hi.. Thank you for the suggestion.

I ran the command on VNX and it shows the foolowing output.

[nasadmin@emc_cs0 ~]$ server_devconfig server_2 -list -scsi -nondisks -all
server_2 :
               Scsi Device Table
name                addr          type  info
gk01                 c0t0l0        disk  1 053107000020
gk161               c16t1l2      disk  5 053115001200
ttape1              c32t0l0      tape  HP Ultrium 5-SCSI Y23U

It identifies the HP Ultrium-5 Tape. Is this correct and sufficient.?

Does it means that VNX is able to communicate with the Tape Drive.?

Hello..

I didi ask support team regarding the NDMP drivers on VNX but they could not give proper answers as of now.

I dont know in which direction to proceed now.

That's showing that the VNX is seeing one drive. But I think you have a library with two. You would generally see three devices, one for the library and one for each drive. You need to specifically load into that one drive the tape(s) you want to use.

Also with Dynamic Drive Sharing between NetWorker and VNX, you need to set the NDMP scsiReserve parameter to 0 on the VNX file. This is of course if you only want to use the library for backup/restore with VNX NDMP.

Unfortunately extended support for these models were dropped at least 12 months ago so I doubt Dell will be of much help. You need to go to the VNX community and post your questions, there may be some people able to help you more. Unfortunately I lost all my documentation.