Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations IamaSherpa on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Blue Screen of Death

Status
Not open for further replies.

labninja

MIS
Jan 18, 2000
44
ZA
I have a SBS server (4.5) that keep on crashing randomly. It started off with 0x0000001a memory dumps. Then I stopped the Backup and Anti-Virus services and now it is coming up with STOP 0x0000001E KMODE_EXEPTION NOT HANDLED blue screens. It is acting as a file, mail and print server. Microsoft advised installing the latest service pack (SP6a). Has anybody tested SP6a on SBS ver 4.5? Any other ideas?
 
I had SBS 4.0 crash randomly too.<br>
but I did not get any blue screens<br>
it just shut off and re-booted.<br>
It as one of two things which I replaced both.<br>
The Power supply for the case or the voltage regulator on the motherboard.<br>
I got a new case and motherboard.<br>
It's been running fine now for 4 months<br>
We have since upgraded to 4.5 and still have no problems.<br>
<br>
<br>

 
I agree with DoupP that it might be a power problem - if you don't have damaged components in the server yet, are you running a REALLY good UPS? How about server room temperature? If you're not chilly when you're in the room with the servers, the room is too warm! this is especially true with Pentium servers - you could get away with warmer rooms in the olden days with slower CPUs. <p>Fred Wagner<br><a href=mailto:frwagne@ci.long-beach.ca.us>frwagne@ci.long-beach.ca.us</a><br><a href= > </a><br>
 
I don't agree about the power supply. The blue screen of death is either caused by a disk problem, or is a memory parity problem. First, have you run scandisk and defraged the drive? I haven't tested the above mentioned configuration, but it {probably} wouldn't hurt to install the SP, may take care of other problems that you aren't even aware of, and there is an uninstall option, so you aren't really risking anything.<br>
I also agree with Fred Wagner about the room temperature, it can really play havoc with a CPU, and definitly can cause drive AND memory errors, as well as damaging very important components. I like the description, &quot;If you aren't chilly, then the rooms too warm&quot;, but to be a bit more accurate the temperature should stay below 70, 65-68 for ideal running conditions. <p>John Vogel<br><a href=mailto:johnvogel@homepage.com>johnvogel@homepage.com</a><br><a href= HomePage</a><br>WebMaster - DataBase Administrator - Programmer<br>
 
As the others have said, BSOD is almost always a hardware problem. Also try removing and re-seating your RAM, expansion card, CPU, cables, etc. Connections can get flaky due to high temp and also due to temp swings.<br>
<br>
Did this just start or has this machine always acted this way? <p> Jeff<br><a href=mailto: masterracker@hotmail.com> masterracker@hotmail.com</a><br><a href= > </a><br>
 
could be someone dicking about with your machine, <br>
<br>
patching the machine over port 139 (ping attack) would be a start (unless you are bhind a firewall) or do as microsoft propose and install a service pack over the top. <br>
<br>
Ping attacks cause the same effect. (tell me if you think i'm just sadistic!) The effect is also known as Ping of death, and a nuke.<br>
<br>
However, you are right to wonder how the service pack works.<br>
<br>
(why 6 updates?) GOSH!<br>
<br>
---<br>
<br>
Karl <p> Karl<br><a href=mailto:mc_karl@yahoo.com>mc_karl@yahoo.com</a><br><a href= > </a><br> ~ ~ ~ ~<br>
K A R L<br>
~ ~ ~ ~
 
I'm not 100% sure NT has a defrag or scan disk<br>
It's self tuning.<br>
It is not under Accessories, System tools like it is in '9x<br>
I didn't see anything in Disk Administrator either<br>
but I run NTFS on everything.<br>
<br>

 
NTFS will fragment like crazy and NT4 has no supplied defrag. Diskeeper is probably the best product to add on for that. There is a Diskeeper Lite that is free and downloadable. Once you install it you access it by right-clicking on a drive in Explorer and choosing Properties&gt;Tools&gt;Defragment Now. <br>
<br>
There is also no Scandisk although the old CHKDSK functionality is there. Problem with that one is that it requires a reboot and can run a LONG time. (Our parent co. did it on a server with about 100GB of disk and it ran over six days. The system cannot boot until it is finished so its pretty brutal.) <p> Jeff<br><a href=mailto: masterracker@hotmail.com> masterracker@hotmail.com</a><br><a href= > </a><br>
 
Just a question.. Typically how long is the machine on line before it blue screens? If the machine has been running for a while, I doubt it is a hardware related issue. Almost all blue screens (except on initial boot) are software related... i.e. memory leaks within applications.<br>
If the server creates a crash dump file called memory.dmp in the c:\winnt directory, try using dumexam to port it to a text file. Then, in that text file search on the Stop exception. If it is a driver.. many times you will find a driver listed that caused the problem. Now, it is not always easy to determine what the driver does, so you will need to do some digging around. If anywhere in the dumpfile you run across the line something to the effect &quot;No more physical memory&quot; or &quot;out of physical memory&quot;.. something along those lines... you have a memory leak. What this means is... some application... driver...etc. is using memory, and then not releasing all of it when the app terminates.<br>
<br>
There are downloads that supposedly clean memory that you can run to release previously allocated physical memory , but I have never tried any of them. A reboot will clear it out.<br>
<br>
If this happens every few weeks or so, and the problem looks like it is memory related, I would suggest scheduled re-boots over the weekend to clean up memory problems... <br>
<br>

 
I have to agree with BruceP on this one. Except on initial boot a BSOD is a memory leak within an application. At this point there would not be a need to assume that someone is ping attacking your server or to go replacing hardware in the machine. Has any of the software on the machine recently been upgraded or any hardware components upgraded that may require a newer driver? You may need to evaluate what services are loading on the machine as well.<br>
<br>
Take a look at Technet. Technet has some good articles describing blue screens but the one that is the most helpful is entitled Key Screen Details to Examine. It is from the NT resource Kit section so there is no Q number but if you do a query using the following including the quotes you should find the article:<br>
<br>
&quot;Key Screen Details to Examine&quot;<br>
<br>
If you are still experiencing problems post the full machine configuration including make and model and the various software installed on it. So that we can get a better understanding of what you are looking. <p> <br><a href=mailto: oneworld@goes.com> oneworld@goes.com</a><br><a href= > </a><br>
 
Thanx for all the responses. I think I have finally solved it. The resolution was a combination of the following: Added a UPS to the server, using a different power socket; replaced the hubs, replaced the server network cable, moved the server away from a radiator and praying to 10 different gods all at once.<br>
<br>
In the end I think in the end it was environmental rather than hardware and/or software specific!<br>
<br>
Take it easy and keep the dream alive!
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top