Windows Server 2003 odd boot behavior

By jasonbowman ·
This is the behavior in a nutshell...

When we boot the system, it goes through the POST normally, RAID controller finds the volumes and continues, then the instant the boot is handed over to the OS, the drives in the RAID begin a steady flashing that last for approx 45 minutes.

The screen stays black during the entire time, and when its done doing whatever its doing, the OS comes up, and we can log in and the server behaves normally for the most part. All the shares and folders are accessible and the server does its duties (DHCP, etc)

There seems to be some odd behavior when we run backups or try to browse the data partition through windows explorer however, nothing fatal, but still troublesome.

If we reboot the server, it repeats the exact same process. These are some of the things we have tried, all to no effect.

We swapped the raid contoller with an identical card, same thing.

We tried the drives in another identical server, they did the same thing in a different machine.

We booted from CD and they do the same thing during the CD boot process, and once you do get to the recovery console, it repeats the 45 minute drive access if you try to run chkdsk.

We put different drives in the server and they behave normally, then when we put the original drives back in, they still do the 45 minute access prior to booting.

There is nothing in the system logs that indicates there is a problem. The server had been running with no problems for months, this behavior began after a normal reboot.

Some tech details...

Dell Poweredge 2900 with 4 drives in a RAID 5 array (3 drives in raid, 1 hot spare) running Win 2K3 Ent. Edition. The RAID coontroller is a PERC 5SI.

I have searched the internet fairly extensively for this issue but have found little that matched these symptoms exactly. I find this odd because this is the second time it has happened to us in the last year, the last time it was a Poweredge 2800 with a PERC 6I contoller.

Neither Dell or Microsoft were able to fix the problem after hours and hours of phone calls, and we are looking at having to reload the domain contoller again.

I am resigned to having to reload the server at this point, but I was curious if anyone else out there has seen this issue and what if anything you can do to fix it or prevent it.

Sorry this ended up being a huge nutshell, thanks for reading and if you can help please share for my sanity's sake.


This conversation is currently closed to new comments.

Thread display: Collapse - | Expand +

All Answers

Share your knowledge

Related Discussions

Related Forums