P
Paul Welsh
Hi all
I have an HP DL380 G4 with 4 GB RAM running Windows 2003 SP2 Standard
running Backup Exec 11 with an attached Quantum autoloader.
The autoloader is connected via an LSI Logic PCI-X Ultra320 SCSI
controller. I see the driver is dated 1/10/02.
All was going well until I added direct attached storage to the server
in the shape of an HP StorageWorks Modular Smart Array 60 containing 5
x 250 GB SATA disks in a RAID 10 configuration (one disk is a hot
spare). The controller for this is a PCI-X HP Smart Array P600 SAS /
SATA Storage Controller. I upgraded the memory at the same time to 4
GB from 1 GB.
The SATA drives are used by other servers to backup around 200 GB of
database files overnight and early each morning via a network share.
The Backup Exec backup runs each night at the same time. It backs up
these database backups, other file servers and an Exchange server.
The daily differential backup is about 500 GB.
The database backups used to be stored on another file server and the
nightly tape backup worked fine. The volume of data hasn't changed,
it's just that the database backups are now stored on the same server
to which the tape drive is attached - the idea was to speed up the
backups of the databases to tape because the network card wouldn't act
as a bottleneck.
Talking of NICs, these are 2 x HP NC7782 cards teamed using HP's
teaming software.
Since the upgrade, the file sharing component of the server hangs on a
regular basis during the overnight/morning jobs (a hard reboot is the
only solution - I can log on but the system's totally unresponsive).
The first time it hung, I got eventID 2021 "The server was unable to
allocate a work item 139 times in the last 60 seconds". I made some
registry changes, as recommended by a Microsoft knowledgebase article
but no joy. I therefore reversed the changes and the problem
continued but with no eventID 2021. On Christmas Day I got Event ID
11, "The driver detected a controller error on \Device\Tape0".
I've worked around the problem by performing a scheduled reboot at
10pm each night. However, there's not much of a window available to
do this at the weekend and last weekend the server hung again, so it's
not capable of going for more than a couple of days without the file
sharing and/or whole server hanging.
Any ideas? I'm considering moving the RAID controller and attached
sotrage to a spare DL380 G4 and using it for the file sharing. If it
hangs then I will know there's a problem with the new hardware. The
other thing I'm considering doing is to move the backup share to the
internal disks in the server that run off an internal RAID controller
(they have been in place for some time). This will achieve a similar
objective of testing the new hardware.
Any further suggestions in troubleshooting this problem appreciated.
I have an HP DL380 G4 with 4 GB RAM running Windows 2003 SP2 Standard
running Backup Exec 11 with an attached Quantum autoloader.
The autoloader is connected via an LSI Logic PCI-X Ultra320 SCSI
controller. I see the driver is dated 1/10/02.
All was going well until I added direct attached storage to the server
in the shape of an HP StorageWorks Modular Smart Array 60 containing 5
x 250 GB SATA disks in a RAID 10 configuration (one disk is a hot
spare). The controller for this is a PCI-X HP Smart Array P600 SAS /
SATA Storage Controller. I upgraded the memory at the same time to 4
GB from 1 GB.
The SATA drives are used by other servers to backup around 200 GB of
database files overnight and early each morning via a network share.
The Backup Exec backup runs each night at the same time. It backs up
these database backups, other file servers and an Exchange server.
The daily differential backup is about 500 GB.
The database backups used to be stored on another file server and the
nightly tape backup worked fine. The volume of data hasn't changed,
it's just that the database backups are now stored on the same server
to which the tape drive is attached - the idea was to speed up the
backups of the databases to tape because the network card wouldn't act
as a bottleneck.
Talking of NICs, these are 2 x HP NC7782 cards teamed using HP's
teaming software.
Since the upgrade, the file sharing component of the server hangs on a
regular basis during the overnight/morning jobs (a hard reboot is the
only solution - I can log on but the system's totally unresponsive).
The first time it hung, I got eventID 2021 "The server was unable to
allocate a work item 139 times in the last 60 seconds". I made some
registry changes, as recommended by a Microsoft knowledgebase article
but no joy. I therefore reversed the changes and the problem
continued but with no eventID 2021. On Christmas Day I got Event ID
11, "The driver detected a controller error on \Device\Tape0".
I've worked around the problem by performing a scheduled reboot at
10pm each night. However, there's not much of a window available to
do this at the weekend and last weekend the server hung again, so it's
not capable of going for more than a couple of days without the file
sharing and/or whole server hanging.
Any ideas? I'm considering moving the RAID controller and attached
sotrage to a spare DL380 G4 and using it for the file sharing. If it
hangs then I will know there's a problem with the new hardware. The
other thing I'm considering doing is to move the backup share to the
internal disks in the server that run off an internal RAID controller
(they have been in place for some time). This will achieve a similar
objective of testing the new hardware.
Any further suggestions in troubleshooting this problem appreciated.