The important part of your message is:
eventid 50 "{Lost Delayed-Write Data} The system was
attempting to transfer file data from buffers to
\Device\HarddiskVolume4. The write operation failed, and only
some of the data may have been written to the file."
This means that you have a problem with your disk drives. This can be a
communication problem (you mentioned changes to the fabric) or a bad disk
giving timeouts due to retries or similar.
My suggestion is that you investigate the hardware first before looking at the
software.
Bye
Ernie
-----Original Message-----
From: Jon Hickmott [mailto:jonny AT UWO DOT CA]
Sent: Friday, April 11, 2003 5:49 PM
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Subject: [Networker] eventid 50 warnings while backing up a W2K cluster
SYSTEM STATE.
Our environment:
* 2 Windows 2000 advanced servers running cluster service
* Shared cluster disks attached via fiber
o Qlogic HBAs
o Brocade silkworm switch
o IBM FAStT700 Storage Array
* Networker version 6.1.3 on client and server.
* Networker server platform is a Sun Solaris 8 box
Our problem
When performing a backup of the SYSTEM STATE of the Windows
2000 client (just the local disks not the cluster disks) that
is currently in charge of the cluster service we start to see
the following warnings.
eventid 50 "{Lost Delayed-Write Data} The system was
attempting to transfer file data from buffers to
\Device\HarddiskVolume4. The write operation failed, and only
some of the data may have been written to the file."
These warning occur in vast numbers and our clients lose
access to their areas on the cluster disks. The cluster
service does not fail. When manually failed over to the other
server all returns to normal for the cluster service.
\Device\HarddiskVolume4 is one of the cluster disks which we
are not backing up. These warnings seem to have something to
do with cache on the disk. We have turned off write cache on
all disks and still the warnings come.
This problem started when we modified the fiber environment.
Before the servers were directly attached to the FAStT700
storage array. Now they are attached through the brocade
switches.
My question
What goes on when NetWorker performs the SYSTEM STATE, SYSTEM
FILES and SYSTEM DB Savesets. Is there a process that may
effect the state of the disks or force out a clearing of
cache? Should we have some specific settings for these fabric
attached disks.
Thanks for any help you can give.
Jon
--
Jon Hickmott - jonny AT uwo DOT ca - NSC211c - (519) 661-2111x86025
Unix/Database administrator - ITS - UWO
"There are only 10 kinds of people in the world --
Those who understand binary, and those who don't."
--
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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
--
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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|