I am having a similar problem with the same error messages.  We added a new fresh install 8.0.0.1 (32-bit) storage node (Windows 2008 Enterprise 32-bit SP2) to the current environment that connects to our IBM 3494 tape library which contains 6 x 3592J1A tape writers.  Previously, the storage node we used for this tape library was working without issue, but it was running Windows 2003 Server which is not supported with NW 8.0.  So I brought up a new server and configured it to control the tape library as a replacement.  Once the configuration was completed, I allocated all of the barcodes that were still in the library so it could use the tapes.  Initially, all seemed fine with the tape library on the new storage node, but while the tape volume is marked as full in the slot listing, the tape writer still thinks it is writing to the tape and will not let it go, i.e., unmount it.

The NW Server is still on 7.6.2.7 for the moment and running on Windows 2008 R2 SP1 64-bit.  All of the IBM drivers for the 3592 tape writers were upgraded and the IBM Automated Tape Library service is up to date, running, and can communicate with the library using "mtlib" commands such as "mtlib -l -qI" which returns the entire inventory of the llibrary.  We hard-code a 256K block size for the tapes per Networker Technical Support recommendations a month or so ago, but the 3592's default to 384K, however, I am under the impression that it supports variable block sizes as well.

The tapes will mount, label, and can be written to in Networker (manually and through a scheduled job).  The problem is that when it is finished with one tape and needs to continue the saveset onto a second tape, the first tape's status never changes to "done writing, full" and it still shows "writing at 50MB/s, 877GB" although nothing is writing to it (the speed it shows varies, but never changes/updates and the size also does not change and reflects the final save size for that portion of the saveset on the tape volume).  I tried unmounting the tape from the drive manually, but it will not let me do that either:

"Automatically terminating operation `OP_UNLOAD', instance 54, on jukebox `rd=tamsbkpnw02:3494'. Cannot allocate the 1 required device(s).";

Storage Node daemon.raw messages:

80042 1345724970 2 0 0 4032 1876 0 tamsbkpnw02.uch.org nsrmmd NSR warning 41 Setting block size for device %s to %d KB 2 21 24 rd=tamsbkpnw02:\\.\Tape2 1 3 256

0 1345724970 1 5 0 4032 1876 0 tamsbkpnw02.uch.org nsrmmd NSR notice 2 %s 1 0 67 tape_cdi_max_recsz -check_cdi_infop: CDI using v_fd for path.handle

89230 1345724970 0 0 0 4032 1876 0 tamsbkpnw02.uch.org nsrmmd NSR info 23 Set %s block size to %d 2 23 9 \\.\Tape2 1 6 262144

89230 1345724970 0 0 0 4032 1876 0 tamsbkpnw02.uch.org nsrmmd NSR info 23 Set %s block size to %d 2 23 9 \\.\Tape2 1 6 262144

89230 1345724970 0 0 0 4032 1876 0 tamsbkpnw02.uch.org nsrmmd NSR info 23 Set %s block size to %d 2 23 9 \\.\Tape2 1 6 262144

89230 1345724970 0 0 0 4032 1876 0 tamsbkpnw02.uch.org nsrmmd NSR info 23 Set %s block size to %d 2 23 9 \\.\Tape2 1 6 262144

80042 1345725010 2 0 0 3584 1120 0 tamsbkpnw02.uch.org nsrmmd NSR warning 41 Setting block size for device %s to %d KB 2 21 24 rd=tamsbkpnw02:\\.\Tape0 1 3 256

0 1345725010 1 5 0 3584 1120 0 tamsbkpnw02.uch.org nsrmmd NSR notice 2 %s 1 0 67 tape_cdi_max_recsz -check_cdi_infop: CDI using v_fd for path.handle

89230 1345725010 0 0 0 3584 1120 0 tamsbkpnw02.uch.org nsrmmd NSR info 23 Set %s block size to %d 2 23 9 \\.\Tape0 1 6 262144

89230 1345725015 0 0 0 3584 1120 0 tamsbkpnw02.uch.org nsrmmd NSR info 23 Set %s block size to %d 2 23 9 \\.\Tape0 1 6 262144

89230 1345725015 0 0 0 3584 1120 0 tamsbkpnw02.uch.org nsrmmd NSR info 23 Set %s block size to %d 2 23 9 \\.\Tape0 1 6 262144

NW Server daemon.raw messages (when trying to unmount tapes manually while full, but status of tape drive shows it as "writing":

12363 1345728836 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 55 [Jukebox `%s', operation %s]. Initiated operation `%s'. 3 0 19 rd=tamsbkpnw02:3494 0 2 52 0 67 operation: "Unload"; operation device: "rd=tamsbkpnw02:\\.\Tape5";

12363 1345728882 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 55 [Jukebox `%s', operation %s]. Initiated operation `%s'. 3 0 19 rd=tamsbkpnw02:3494 0 2 53 0 67 operation: "Unload"; operation device: "rd=tamsbkpnw02:\\.\Tape4";

12363 1345728890 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 55 [Jukebox `%s', operation %s]. Initiated operation `%s'. 3 0 19 rd=tamsbkpnw02:3494 0 2 54 0 67 operation: "Unload"; operation device: "rd=tamsbkpnw02:\\.\Tape1";

7224 1345728899 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 5 %s %s 2 0 48 [Jukebox `rd=tamsbkpnw02:3494', operation # 52]. 0 137 Automatically terminating operation `OP_UNLOAD', instance 52, on jukebox `rd=tamsbkpnw02:3494'. Cannot allocate the 1 required device(s).

12361 1345728899 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 27 %s Finished with status: %s 2 0 48 [Jukebox `rd=tamsbkpnw02:3494', operation # 52]. 0 6 failed

7224 1345728943 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 5 %s %s 2 0 48 [Jukebox `rd=tamsbkpnw02:3494', operation # 53]. 0 137 Automatically terminating operation `OP_UNLOAD', instance 53, on jukebox `rd=tamsbkpnw02:3494'. Cannot allocate the 1 required device(s).

12361 1345728943 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 27 %s Finished with status: %s 2 0 48 [Jukebox `rd=tamsbkpnw02:3494', operation # 53]. 0 6 failed

7224 1345728953 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 5 %s %s 2 0 48 [Jukebox `rd=tamsbkpnw02:3494', operation # 54]. 0 137 Automatically terminating operation `OP_UNLOAD', instance 54, on jukebox `rd=tamsbkpnw02:3494'. Cannot allocate the 1 required device(s).

12361 1345728953 2 0 0 2892 2888 0 emcnetworker1.uch.org nsrd 27 %s Finished with status: %s 2 0 48 [Jukebox `rd=tamsbkpnw02:3494', operation # 54]. 0 6 failed

Any thoughts?  8.0 storage nodes are supposed to be compatible with 7.6.2.7 NW Servers.  Thanks in advance for your help!

bsock79 wrote: Any thoughts?  8.0 storage nodes are supposed to be compatible with 7.6.2.7 NW Servers.  Thanks in advance for your help!

Who said that?

Hi Hrvoje

Yep they are according to NW8 Install Guide:-

Interoperability and backward compatibility considerations

This section provides important information about the interoperability and backwards

compatibility of the NetWorker 8.0 software.

Before updating any machine in the datazone to NetWorker 8.0, review these

interoperability considerations:

◆ A NetWorker 8.0 server is not compatible with a NetWorker 7.6.x or earlier storage

node.

Update all of the NetWorker storage nodes to version 8.0 before updating the

NetWorker server.

NetWorker 7.6.x and earlier storage nodes are disabled by a NetWorker 8.0 server.

◆ A NetWorker 7.5.x server cannot be directly updated to NetWorker 8.0.

Update the NetWorker 7.5.x datazone to version 8.0 in the following order:

• Update the NetWorker 7.5.x server to version 7.6.x.

• Update each NetWorker storage node to version 8.0.

• Update the NetWorker 7.6.x server software to version 8.0.

• Update the NetWorker clients to version 8.0.

◆ A NetWorker 7.5.x Console server that is not also the NetWorker server can be directly

updated to the NetWorker 8.0 software.

◆ When the Console server is not the NetWorker server, update the Console server to

version 8.0 before updating the NetWorker server.

◆ A NetWorker 8.0 storage node is compatible with a NetWorker 7.6.x server.

◆ A NetWorker 8.0 client is compatible with a NetWorker 7.5.x and 7.6.x server.

◆ NetWorker 7.5.x and 7.6.x clients are compatible with a NetWorker 8.0 server.

◆ If the host operating system of a NetWorker server, a storage node, or a client will also

be updated, update the operating system first, then update the NetWorker software.

◆ NetWorker Module for Microsoft Applications (NMM) 2.3 is not supported with

NetWorker 8.0. Update the NMM software to NMM 2.4 before updating the NetWorker

server and storage nodes to NetWorker 8.0.

I doubt this works with 7.6.2.x... perhaps later 7.6.3.x or 7.6.4.x, but not 7.6.2.x. When you have release which has some many changes, you don't wish to make partial upgrade (like sn yes and bs no) as that is always asking for trouble. Unless you have test environment and time to play around.

When you say you tried to manually set teh block size I assume you tried by setting the enviroment variable on the Storage Node server? and in the NMC?

I have used a command similar to this in the past:

NSR_DEV_BLOCK_SIZE_LTO_ULTRIUM_3=128

NSR_DEV_BLOCK_SIZE_LTO_ULTRIUM_4=128

I then set the block size in the Networker NMC GUI to match.

Hope this helps.

Hi eivinda

which method you change the env var ?;

◆ “Device settings in the NetWorker Administration interface” on page 187 (NO service restart)

◆ “Device settings as environment variables” on page 188 (must restart service)

Hi Hrvoje

Just copy paste as what's in doco... but I would not do that in my env though the doco says it supports it..

I'm having weird block size issues using RHEL v5.8 (32bit) with a Quantum Scalar i80 library and HP LTO5 drives.

I have tried setting various different block sizes both in the NetWorker GUI and globally using the nsrrc script.

The majority of the problems I'm seeing is during tape label operations but occasionally I will get an I/O error and NetWorker will make the tape as full.

Anyone know of a solution or workaround?

-Brian

Our problem was a faulty drive.

Not sure if I should make this discussion answered as it has lots of other ppls questions in it now.

E;>

We have the same problem with setting block size. Did you get this sorted out?

Our setup is

• Solaris 10 on SPARC
• Networker 8.0.0.3
• IBM 3584 with LTO3 drives

It worked when we upgraded from 7.6.2.5 to 8.0.0.3 but suddenly it started to have these errors.

Karl Dag