It may not have anything to do with the cluster but the
cluster dies or both nodes are effected so just a
thought....
Configuration
SQL 7 Ent with service pack 4 and MS 03-31
Win2k with SP 3
8 CPU
8 GIG memory with /3GB and /PAE Switch
Errors
1450 Insufficient resources
Error 823
BackupIoRequest::WaitForIoCompletion: write failure on
backup device 'J:\SQLDATA\WebInfo_Log.LDF'. Operating
system error 1450(Insufficient system resources exist to
complete the requested service.).

Any suggestions?

Looks like you ran out of disk space since it's a write
failure, or you have a bad disk.

There is plenty of disk space. The error "Operating system
error 1450(Insufficient system resources exist to complete
the requested service)." is the key?

Disk space on drive J:\? Are you sure?

If you are seeing those types of errors, generally, it's
an I/O based error. The 823 error is very telling; 1450
may be a symptom of it.

Have you checked your error log for disk failures and such?

As per SQL Server 2000 Books Online topic "Error 823"

Error 823
Severity Level 24
Message Text
I/O error %ls detected during %S_MSG at offset %#016I64x in file '%ls'.

Explanation
Microsoft® SQL Server™ encountered an I/O error on a read or write request
made to a device. This error usually indicates disk problems. However,
additional kernel messages in the error log, recorded before error 823,
should indicate which device is involved.

Action
Check the accessibility and condition of the device in question.

Run hardware diagnostics and correct problems, if possible.

Restore damaged files from the latest database backup. Restoring from a
database backup should always be considered the primary means of fixing a
damaged database.

If you don't have a backup or if the errors detected are very isolated, the
repair functionality of DBCC CHECKDB may be useful. However, using DBCC
CHECKDB can be more time consuming than restoring the damaged files from a
backup, and you may not be able to recover all your data .

Caution If running DBCC CHECKDB with one of the repair clauses does not
correct the problem or if you are unsure how this process may affect your
data, contact your primary support provider.


HTH,

Uttam Parui
SQL Server Developer Support Engineer
Product Support Services
Microsoft Corporation

This posting is provided "AS IS" with no warranties, and confers no rights.

Are you secure? For information about the Strategic Technology Protection
Program and to order your FREE Security Tool Kit, please visit
http://www.microsoft.com/security.

This cluster is attached to a SAN. The error is on all
drives so its not hardware based as there are multiple
disk controllers(unless they all go down for a short
period of time)? Also, after the reboot, the servers last
for about 4, 5 days? The only direction my thoughts go is
that its a server level (kernal memory) issue. Any ideas
about SYSTEM PTE?

How can you say that the error is not HW based? If you're
seeing the problem on all drives, it is definitely HW
based. Having multiple controllers does not mean that you
do not have problems. Could be anything from bad
controllers, bad disks, driver issues, etc. SANs are
reliable, but one of the biggest problems that places
occur is the SAN. Again, the 823 is glaringly obvious.'

You really need to go through EVent Log and probably open
a case with PSS to help you resolve this.

Sorry if I have given an impression that i am saying
anything with authority. PSS is working on this issue and
they have all cluster logs, perfmon, HW reports etc etc.
Its 2 weeks and production cluster goes DEAD is really
scary?

Looks as if removing /3GB switch did the trick?

Sometimes /3GB and /PAE do not play well, but it also
depends on many other things. If that solved your
problem, that's great. My recommendation is to always
pick one memory model and stick to it. Not much past 8 GB
anyway, you can't combine /3GB and /PAE.