Hello,
We have been receiving errors from NetWorker regarding the block size
(see error message below) when recovering data from tapes on our
StorageTek L80 LTO tape library. We were trying to recover data from a
Unix host, not an NT box. This library is connected to a Linux
storagenode server running NetWorker 6.1.1. This is our only
storagenode. We are not using DDS, and the few NT clients that we have
are only clients. We have no NT NetWorker servers or storagenode
servers. The primary server is running 6.1.1. also but under Solaris.
The recoveries work, but they seem to take longer than they should,
particularly given the fact that we were just recovering one file, and
sometimes we've had to retry after receiving on one occasion a "tape
busy" message. One suggestion from this forum was to check the block
size definition as discussed in the admin guide and perhaps set the
environment variable(s) as necessary. I read the admin guide on this,
but there still is the question of what variables to use and WHAT value
to set these variable(s) to?!
For example, in the guide it mentions:
A number of environmental variables can be used to adjust several
NetWorker device parameters:
Environment variables should be set in the NetWorker startup script
before the lines: (echo 'starting NetWorker daemons:') > /dev/console
Two variables that looked promising are:
NSR_DEV_BLOCK_SIZE_DEVICE_NAME is organized in units of KB. The value
set must be a multiple of 32, with a minimum value of 32. Maximums are
determined by platform, SCSI driver, and device. For example:
NSR_DEV_BLOCK_SIZE_4MM_20GB=64
NSR_DEV_TAPE_FILE_SIZE_DEVICE_NAME is organized in units of
NSR_DEV_BLOCK_SIZE_DEVICE_NAME and is the number of blocks written
between filemarks. These filemarks are used to locate a particular spot
on the tape during recovery, and more filemarks will generally lead to
faster
positioning. For example: NSR_DEV_TAPE_FILE_SIZE_TZ89=512
? On UNIX platforms, the NetWorker software writes a filemark by closing
and reopening the tape device; requiring a second or two. If this value
is too small, throughput may be slowed.
Should I add both these variable to the startup script on our
storagenode server and if so, what values should I use?
Here are the types of error messages we've been seeing:
"rd=snode:/dev/nst2": Cannot decode block. Verify the device
configuration. Tape positioning by record is disabled.
Thanks.
George Sinclair
George.Sinclair AT noaa DOT gov
--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
