We have configured it on Fibre switch. Although the library is SCSI but we have a MBR connected. NW 6 is in use.
Things have been working good for years but for some testing purposes we had shut down this server and library and connected the SCSI library directly to a new backup server. We installed NW 7.3.2 there to configure NDMP backups.
When we used the library on the new setup, it would correctly perform all the operations but everytime it would give an error - INIT ELEM STS RNG failed error 5.
I configured one of the drives for NDMP - remote configuration but when we use this drive for labelling, it stops at "Verify label in progress" for almost 9-10 min. till I forced stop the action & unmounted the tape.
We again reconnected the library to the original configuration. Now everything works fine there but everytime we issue a media related command, the library checks the physical inventory of the entire library.
Even when we start a group, it checks the complete inventory(only physically - does not load any volume) and the it loads the correct volume and takes the backup normally.
Any ideas. We have logged a case with the hardware vendor but we have done any reconfig. on Networker so Networker should not be a problem.
NDMP license has (dedicated?) storage node license integrated so you don't need it.
If you have device used for filer in DDS it means that filer will write data via SAN to tape device and other clients can write it via network to some other storage node where this same device is presented (eg. backup server).
This thing with inventory is something I have seen once or twice in the past and I think it was also some MSL library... not sure as I don't remember that long (it must have been at least 4 years I believe), but I think vendor did several changes like library and rive fw update and perhaps something else and at the end something from those changes did make a difference.
I am using 732 Jumbo 11 only. I am very particular about using the best possible version especially when testing on site. It otherwise creates a wrong impression.
Just wanted to check if my config. steps were ok -
1. I ran jbconfig on NW server since robotics is controlled by NW server. 2. While configuring drives, I marked one of them for use of NDMP backups and mentioned drive path as NDMP_Server:\\.\tape0 (\\.\tape0 as visible on NW Server) 3. I have created a group, schedule & pool separately but actually don't know if there is any place to mark NDMP. 4. I have created a client with filer's name and marked NDMP in that apart from giving the ndm specific backup command etc.
I think that's it, but I'm puzzled by device... why NDMP_Server:\\.\Tape0? Usually NDMP filer has it's own device naming convention, but I have never heard of \\.\Tape0 on filer. What filer is that?
ble1
6 Operator
•
14354 Posts
•
56186 Points
576
0
Posted January 13th, 2007 09:00
If you have device used for filer in DDS it means that filer will write data via SAN to tape device and other clients can write it via network to some other storage node where this same device is presented (eg. backup server).