Networker

[Networker] Update w2003 problems with tape full

2006-08-08 05:52:26
Subject: [Networker] Update w2003 problems with tape full
From: Maarten Boot <maarten.boot AT MBU DOT HR>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Tue, 8 Aug 2006 11:46:21 +0200
Dear List,

 

Context: Win2003, Networker 7.3.1, San , 3 extra storagenodes Hpux + Win
2003, StorageTek SL500 with 2x LTO-3

 

Tape full problems are over, we found 2 possible candidates.

 

1)       RSM was running on one storagenode (win2003 cluster node)

2)       One of the san switches has intermittent errors in the log (and
a older firmware as the other) we upgraded the firmware and the errors
have gone away. 

 

We also found that 3 extra machines who are not storage nodes were also
in the zone that shares the Library and the drives to the storage nodes
they have been removed 

 

The staging problems we had turned out to be related to the list if
devices inside networkers resource file. The last one added (our HP box
with currently only 100Mhz net was the latest in added abd therefore the
first candidate for any operation. We plan to upgrade to a 1Gb interface
but as this is production we have to plan this. Setting the read
storagenode on the library solved that.

 

Currently I have removed alll storage nodes from the servver and we had
last week no problems related to this anymore. I will add them slowly
this weeks.

 

 

Other problems have turend up regarding automatic mounting and cancelled
clone scripts. 

They seem to stop automatic loading of volumes. 

One remedy against that is to stop the services on the windows backup
server and then rename the ."c:/Program files/Legato/nsr/tmp" folder to
tmp-old. 

After restarting the services any old memories seem to disapear and the
server responds as normal again. 

 

I used to do this on my unix networker machine after the state of the
library was confused it seems to be possible stilll with 7.3.1

 

So far,

 

Maarten

 

________________________________

From: Dave Gold - News [mailto:dave2 AT cambridgecomputer DOT com] 
Sent: Saturday, August 05, 2006 3:11 AM
To: Maarten Boot
Subject: nsr mailing list: followup win2003

 

Hi Maarten,

 

I'm catching up on a few of mailing list stuff, so hopefully this has
come up already.

 

First, have you put the Windows 2003 "TUR" registry change on so Windows
doesn't send SCSI TUR checks (via plug and play service) every second or
two? That could cause tape rewinds unexpectedly. (Search for TUR as
support.microsoft.com for directions. i suggest making the change based
on the tape driver)

 

Second, if HP, disable SCSI Info and Fibre Array in the hp management
agents.

 

Third, if on a SAN, check the switch ports and hba's for physical
errors. Tape seems to get these more often than disk, and the solution
is usually:

--Make sure to use the hba vendors drivers, not microsoft one.

--Install the hba vendors utilities so you can check for errors.

--Hard code switch speed and device or initiator. (I normally do a
bridge and switch port first, since the bridges have the most problems)

 

Encoding errors, things like that are physical errors--FC says 1 error
every 7 or 11 errors, so if you get errors at the physical layer, that
is critical to fix.

 

And of course turn off removable storage service, and disable it...but
you knew that one :)

 

If this is useful, synopsis to mailing list please!

 

Yours,

 

Dave


To sign off this list, send email to listserv AT listserv.temple DOT edu and 
type "signoff networker" in the
body of the email. Please write to networker-request AT listserv.temple DOT edu 
if you have any problems
wit this list. You can access the archives at 
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER

<Prev in Thread] Current Thread [Next in Thread>
  • [Networker] Update w2003 problems with tape full, Maarten Boot <=