Start a Conversation

Solved!

Go to Solution

1 Rookie

 • 

82 Posts

2858

October 7th, 2019 06:00

Tapes start to stuck (cant unload)

Hi, One week ago when montly backups(to tape) started, the emc networker began to create bottle necks  with alerts like  writing process was idle or  waiting for specific volume.

Today when i check weekly backups(to tape) same issue and i cant unload the tapes with GUI.

 

Any idea?

clipboard_image_0.png

 

We use NMC 91.1.3 Build 189 (on Windows)

 

Thanks in Advance.

1 Rookie

 • 

82 Posts

October 11th, 2019 07:00

  1. I going to look into the logs. Should I to pay attention to something in particular?
    • I do not know as I do not know all the details. Just look carefully whatever looks strange.
  2.  Yesterday I founded and  stopped a clone session that was  stuck from the past weekend (2 days for 17 GB)
    • The logs:
      • Unable to start save session with server networker for exchangedag:: Failed to connect to storage node on networker: nsrrecopy failed to authenticate with nsrmmd on host paipoteres.enami.cl: Timed out
      • How do i extend the leasing time to avoid the time out?
    • I would not look at a timeout but check why the connection failed - obviously for a longer period. This looks more like a 'network related issue'.
      • network tests are ok, The library and the server are in the same vlan even on the same rack.
      • Today I restarted the networker server and show up another process stuck for 7 days. That message wasnt there before the restart.
  3. I enter  directly to the powervault tape library GUI and unmounted the tapes.
    • i try Cleaning proccess  .and takes to long 
    • Please be careful when dealing with your action verbs - 'mounting' is "establishing a data channel between two parties" (here: NW & the tape drive after a tape has been loaded). The jukebox usually does not know which software it is connected to. So all what you can do from the jukebox is 'unloading' or 'ejecting' the tape from the drive. Sorry for being so picky but using the right wordings will help to understand the situation better.
      • Thanks, thats helps me a lot to understand better ( the language barrier is an issue for  technical understanding)
    • As a cleaning tape does not contain any data, such process usually just only takes a few seconds (inserting the 'tape', trying to determine whether it can be read at all, eject). I wonder what you mean by "it took to long".
      • It's took to long: means hours.  After the networker server restars cleaning action messages showup OK.
      •  
      • clipboard_image_0.png
    • Cleaning is usually over-estimated. Over the seven years I used out library with 14 LTO-5 drives I never had to run a cleaning job.

After the suggestions  i opened a ticket with  our vendor. Thanks in advance.

4 Operator

 • 

1.3K Posts

October 7th, 2019 07:00

You can try an unmount the tape outside NetWorker using the sjimm command, The command is as follows. If the unload still fails then you will need to get the tape drive looked at by the vendor.

sjimm jukebox drive x slot x

eg: sjimm scsidev@0.4.0 drive 4 slot 1

You can get the jukebox scsi address from the inquire output and the drive and slot numbers using sjirdtag command.

 

 

2.4K Posts

October 7th, 2019 08:00

May I suggest that you follow this procedure:

1. Carefully investigate the reason why the 'waiting' situation occured and how to solve it.

2. Try to stop the pending jobs if they still exist.

3. Manually unmount/unload the tapes, if still necessary.

Otherwise you most likely will sooner or later re-encounter the 'hanging' scenario again.

 

1 Rookie

 • 

82 Posts

October 7th, 2019 08:00

Hi thanks for answering,  i am going to lookup that command  and reply later.

1 Rookie

 • 

82 Posts

October 8th, 2019 06:00

Hi bingo (again)

  1. I going to look into the logs. Should I to pay attention to something in particular?
  2.  Yesterday I founded and  stopped a clone session that was  stuck from the past weekend (2 days for 17 GB)
    • The logs:
      • Unable to start save session with server networker for exchangedag:: Failed to connect to storage node on networker: nsrrecopy failed to authenticate with nsrmmd on host paipoteres.enami.cl: Timed out
      • How do i extend the leasing time to avoid the time out?
    •  
  3. I enter  directly to the powervault tape library GUI and unmounted the tapes.
    • i try Cleaning proccess  .and takes to long 

2.4K Posts

October 8th, 2019 07:00

  1. I going to look into the logs. Should I to pay attention to something in particular?
    • I do not know as I do not know all the details. Just look carefully whatever looks strange.
  2.  Yesterday I founded and  stopped a clone session that was  stuck from the past weekend (2 days for 17 GB)
    • The logs:
      • Unable to start save session with server networker for exchangedag:: Failed to connect to storage node on networker: nsrrecopy failed to authenticate with nsrmmd on host paipoteres.enami.cl: Timed out
      • How do i extend the leasing time to avoid the time out?
    • I would not look at a timeout but check why the connection failed - obviously for a longer period. This looks more like a 'network related issue'.
  3. I enter  directly to the powervault tape library GUI and unmounted the tapes.
    • i try Cleaning proccess  .and takes to long 
    • Please be careful when dealing with your action verbs - 'mounting' is "establishing a data channel between two parties" (here: NW & the tape drive after a tape has been loaded). The jukebox usually does not know which software it is connected to. So all what you can do from the jukebox is 'unloading' or 'ejecting' the tape from the drive. Sorry for being so picky but using the right wordings will help to understand the situation better.
    • As a cleaning tape does not contain any data, such process usually just only takes a few seconds (inserting the 'tape', trying to determine whether it can be read at all, eject). I wonder what you mean by "it took to long".
    • Cleaning is usually over-estimated. Over the seven years I used out library with 14 LTO-5 drives I never had to run a cleaning job.

4 Operator

 • 

1.3K Posts

October 8th, 2019 22:00

Also, if the tape drive is unavailable and disabled due to exceeding amount of errors the backup server will not be able to connect to the nsrmmd and might cause this error as well. I would suggest you get the tape library/tape drive checked by the respective vendor instead of just digging. around. Once you eliminate the tape library(I doubt that its is not the culprit here) then you can look at other components.

1 Rookie

 • 

82 Posts

October 11th, 2019 07:00

Hi, thanks for asnwering , After the suggestions  i opened a ticket with  our vendor. Thanks in advance.

No Events found!

Top