Hallo Shelley,
please do not compare NDMP and linux, this does not work because on linux
the networker software is used but on the netapp this is replaced by the
ndmp software supplied by NetApp.
Unfortunatly the loadsleep timer is a global value, that means that if the
library is shared then setting the timer will affect all systems.
In the debug information, that is also logged on the netapp in the ndpmd
logfile, you must keep your eyes open for messages that do not return
NMDP_NO_ERR, such as NDMP_IO_ERR, if you look at the message type then you
can an idea what is going wrong. A suspect combination is NDMP_TAPE_MTIO
followed by NDMP_IO_ERR, which could mean that a rewind is send to the tape
drive but the drive is not ready.
Bye
Ernie
-----Original Message-----
From: Shelley L. Shostak [mailto:sls AT qstech DOT com]
Sent: Friday, July 04, 2003 1:52 AM
To: Bokkelkamp Ernst
Cc: Legato NetWorker discussion
Subject: RE: [Networker] Summary: Netapp NDMP backups with NW7.0
Thanks for your suggestions, Ernie. I think I have resolved most of the
issues. I can read from the tapes (except for one instance where I have two
backups in a single day; then networker fails!) I find it difficult to
believe that taking a working networker 6.1.3 server and upgrading to 7.0
would cause hardware problems to manifest themselves. I did crank up
debugging to 70, but it did not really reveal anything meaningful to me. As
for the load/unload timers, I have NO problems with the two drives that are
connected to the linux box. I believe this to be a networker related issue.
I am going to keep my fingers crossed and hold my breath.
Thanks for your great suggestions!
Shelley
On Wed, 2 Jul 2003, Bokkelkamp Ernst wrote:
> Date: Wed, 2 Jul 2003 13:00:22 -0700
> From: Bokkelkamp Ernst <ernst.bokkelkamp AT siemens DOT com>
> To: Legato NetWorker discussion <NETWORKER AT LISTMAIL.TEMPLE DOT EDU>,
> Shelley L. Shostak <sls AT QSTECH DOT COM>
> Subject: RE: [Networker] Summary: Netapp NDMP backups with NW7.0
>
> First thing you should do is to get the NDMP specification and try to
> understand it. Even if you do not understand it completely, it will help
you
> with the next step.
>
> Next activate debug on the filer with something like "ndmpd debug 70"
(don't
> ask me about the 70, I don't know)
>
> Then INCREASE your load sleep timer to something like 150 second or more.
> If you do not do this then you will see some interesting messages on the
> filer ;-) because networker might be trying to acces the tape to early.
Btw.
> Networker gets a completion as soon as the hand has mounted the tape,
while
> the drive is still busy.
>
> If this does not work, then the most difficult part follows: Analyze your
> SCSI configuration, try dump on the filer.
> Are you sure that the cabling, HBA, tape drive AND terminator match, is
your
> cabling within maximum allowed length ?
>
> Bye
> Ernie
--
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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
