• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 327
  • Last Modified:

Excessive hard drive activity on Win2003 server.

I noticed that my Win2003 server is accessing the hard drive constantly. I’m using it as a web server and noticed it when I had to change the power supply out. Then I turned it back on it just started accessing drive like crazy. It doesn’t happen when I first turn it on.  I have looked for anything out of the ordinary but can’t seem to track the problem down. Anyone have any ideas?

http://www.gettinspooky.com/temp/disk.jpg
0
caldwels1895
Asked:
caldwels1895
  • 5
  • 3
  • 2
  • +1
1 Solution
 
InteraXCommented:
Hi caldwels1895,

The only thing going on in that screen shot is 24K of writes to the C drive in 1 second. Nothing to really be worried about depending on the sites hosted and the activity expected/RAM.

Apart from the disk access, does anything else seem odd? What about disk I/O over a period of time?

Good Luck,

Chris
0
 
caldwels1895Author Commented:
I'm watching the hard drive light on the front of it and it's going nuts.
0
Industry Leaders: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

 
InteraXCommented:
caldwels1895,

Try running performance monitor from under Administraive Tools. This should identify all activity that the OS is aware of. From what you've shown, there doesn't appear to be much going on. Does the HDD sound as if it's being accessed?
0
 
Jay_Jay70Commented:
if your disk is going nuts then it can be a tell tale sign of a dying disk, make sure you have a backup ready for if this drive fails....
0
 
caldwels1895Author Commented:
Now it is not doing it! I can't figure out what is happening. I would think it wouldn’t be the drive going dead because I’m not getting the usual errors. I checked the performance monitor and all I can see when it is happening is the disk be access like crazy.
0
 
Chris GralikeSpecialistCommented:
Did you allready consider using RMON to monitor disk I/O and responcible processes. Next to that, if it is hosted on an array you might want to consider checking the logs of the raid controller for anything out of the ordinairy, smart errors etc...

Regards,
0
 
caldwels1895Author Commented:
LOL... I would but this is just a regular PC with Windows 2003 installed. I'm starting to get many hits on the website and maybe its just trying to keep up.
0
 
Chris GralikeSpecialistCommented:
lol :)

then a few questions remain,

1. Do you have a database running on that machine? if so,
    are the transaction logs, database files on the same partition of the HDD in question?
2. Where is the pagefile located? (swap in linux)
3. Is the installation of windows2k3 also on the same HDD?
4. Does the installation run any additional roles like DNS / DHCP / DC / Fileserver / Print spooler etc?

-Regards,
0
 
caldwels1895Author Commented:
1. I'm running an ASP website and it does access a database and they are on a different partition but on the same drive.
2. Not sure
3. Not sure
4. Also runs Print server, DNS, DHCP, and Fileserver.

You did just give me an idea. I did install a printer on it the other day! DUH! I'm going to check and see if the printer software is screwing with me.
0
 
Chris GralikeSpecialistCommented:
No most likely its the database ;-)

Normaly the database disk setup looks like this

C:\ Raid 5             (System Files)          (Distributed parity)
E:\ Raid 10 / 1           (Transaction log) (striped mirror / mirror)
F:\ Raid 5             (Data files)               (Distributed parity)

Usually these are devided over 2 fysical disks defined in the raid configuration. Raid 10 for the transaction log because of the extensive IO generated by the database. Almost all transactions made by the application are handled through the database transaction log. This is more then only the query for data. Raid 5 for the data files and OS because its less expensive then 10 that either loses 1 disk in minimal config for raid 1 and minimal 2 disks for raid 10.

Now you dont have raid, and run it all on one fysical hard drive. This will indeed cause allot of IO going to that one disk. Next to that for most DNS queries, DHCP tasks will most likly also generate IO for that disks. Im not very suprised the disk is very bussy as you mention...

Removing the database and or relocating it on a different system will most likly help allot..

-Regards,
0

Featured Post

Become an Android App Developer

Ready to kick start your career in 2018? Learn how to build an Android app in January’s Course of the Month and open the door to new opportunities.

  • 5
  • 3
  • 2
  • +1
Tackle projects and never again get stuck behind a technical roadblock.
Join Now