Proliant 1600 wont boot (new thread)

[apache_1]

Hi John,

I am starting this new thread as discussed. I am running that window diagnostic tool right now and will post the results..

[apache_1]

Hi John, I ran an Unattended test (1 loop, quick test).. No errors.. Should I run a complete test? More loops etc. (i set the option to stop on first error.)

[JohnWRuffo · Accepted solution]

Yes... work the system out. Full suite of tests and loop it for 10 runs at least (that should be good until the morning)

Insert a CD-ROM in the drive (for testing)and a diskette to record the log to.

[apache_1]

Hi John, I just ran a complete test.. only took like 5 minutes, no errors.. I am going to do as you suggested and write back to you tommorow and we will go from there..

Thanks a lot for your help, see ya tommorow..

[apache_1]

HI John,

AAARRRGH! Just before leaving to run an errand, I took a quick look at the machine and it was down. it said "replace the disk and hit any key etc. etc." so it went down and was trying to come up but could not becuase of the floppy and cd in the drives. Any ways, i removed the cd and floppy and hit enter and here is what it gave me..

172-2 IRC config invalid
-COM port invalid
-PCI com port
Incorrect IRQ

My question is why did it not do it the first or the second time around.. i guess after i put some load on it like you suggested - this happened.. so the machine is not at all stable..

I have to leave now but i'll try to check the forum from home and start on it again tommorow.. any to dos for tommorow morning?

[JohnWRuffo]

172-2 IRC Configuration Invalid means there was conflict with the video controller. The System Inspect/Diagnostics seemed to detect a video card in a PCI slot on the secondary bus.
The "fix" stated in the documentation is to move video controller to a PCI slot on the primary PCI bus.
I am fairly sure you didn't install a PCI video card did you?
Check SW1 and make sure all the switches are OFF except S3.

The first error coupled with these other errors are as if the system configuration was invalid or that SW1 S6 was on? Did you remember to save the System Configuration after the last DIMM was installed?

You had a NMI error logged on Socket 1... I wonder if the root cause was the first DIMM?

[apache_1]

Hi John, Good morning..

You are right, I did not install any video card, and I checked the switches too- they are allright..
Also, I did save the config after every step..

May be its the dimm in the first slot... my question is how come it ran okay all this time.. and how does it fail so peculiarly, and ever so subtly and more importantly - only intermittantly.. is it a regular issue experienced by many? or is this unheardof..

So I should pull out the first dimm and run like 20 loops of testing again and see how it takes it...right?

O guru! too many questions for a tuesday morning but enlighten me O great one, and I'll sing paens of thy glory!! :-)

[JohnWRuffo]

No... not unheard of. RAM can be fickle like this.
The errors you got from that diagnostic test are so off the wall from what I would have expected as the failure point. I would be inclined to run the test again just to discount the test results.

Re-run the SCU again and set the box for diags. This time, set it for a full suite of tests logging errors but do not stop on error only one loop.

But first, Can you send me a survey of your system. You can use that Diagnostics software or you can download and make this diskette:
http://h18000.www1.hp.com/support/files/server/us/download/18396.html

Save the output to a text file and post it as an attachment.

Thanks.

[apache_1]

Hi John, I tried the whole day yesterday to post.. no luck, i think the forum was down..anyways, i am just testing this one out.. will write soon if this goes thru'

[JohnWRuffo]

ok... I'm here for ya.

[apache_1]

Oh! its posting again! here is the low down on the last couple of days.. the server boots up now and then and fails under load or some times fails just like that.. some times it runs okay for some time.. overall, its not stable at all.

but it does not show any errors on the error tab of the utility that you gave me.. however, i have had different error messages during boot up and regular funcitoning..

1. everything was fine and suddenly a blue screen with the message " call you hardware vendor for support ** your system has halted**

2. a critical error occured prior to this power up f1 to continue and f10 for system partition utilities..

John, there are so many other things that I am grappling with right now that it is a little frustrating.. I wanted your advise on it. Is there some one that I could give this server to for trouble shooting?? Do you think it will be worthwhile or just buying a new one is the right thing to do.. the reason i am thinking this is because it is just not stable at this point..

How come sometimes the reults are absolutely clean?? its crazy i tell you..

thanks a lot for all the help and guidance that you are providing!!

[JohnWRuffo]

Well... there are Compaq/HP Partners in every part of the world and some of them are Top Notch. (It's where I cut my teeth on Compaq/HP and they made me take tons of tests before I could work on their products) Most of the VAR shops can get this server going after diagnosing and benching the server after a time and replace what components need replacing. But it will cost...

As much as this server seems to hide its faults until you give it a load, I would be hesitant to deploy it with an application until it passed diagnostics for 24 hours under load. That said, you could end up spending a pile of money replacing parts... even if this server is modular.

The business line is this:

Is it cost effective to spend man hours and money on parts to get this server running or would it be a better choice to purchase a new server with the latest technologies and horsepower. You can weigh the cost in frustration also. The winner in that argument is always who has a warranty. When you have a warranty on the system, you just pick up the phone, call an HP Engineer out and say "fix it!"

The ProLiant 1600 is a workhorse of a server, and I admire them dearly; but the new generations are likely the better choice for you.

For all this servers instability could just be one item... (a stick of RAM, the Processor or a PPM, the cage, etc... not known at the momemnt) but finding that problem will take time. What is your time worth?

If I were just around the corner, I would drive over there and lay on my hands of healing...

[apache_1]

That was good! I will take a look at the relevant costs for fixing it as well as buying the newer ones etc. Any special links or just start from hp.com :-)

I so wish if I could mail it to you shipping prepaid and get it back!!

I'll look around here and let you know how it is going.. In the meanwhile, is there anything else that I can do on it to test how stable it is..

Thanks once again !

[apache_1]

My time is not worth that much but its just that ours is a small start up company and I am doing a lot of stuff in different areas so to find out time for this is the trouble. I have distractions galore and phones that have a ringing disorder.. :-(

[apache_1]

latest one in the error is this... when i booted it just now, it said

TRAP 00000002 =====EXCEPTION=============
then a bunch of memory addresses (i think)

tr=0028 cr0=800000011 cr2=00000000
trapdis etc.etc.

???

[JohnWRuffo]

Here is what I want you to do. Set your crash dump to 64K and post that file here.
Here is the KB File to read:
http://support.microsoft.com/default.aspx?scid=kb;en-us;315263&Product=win2000

[apache_1]

okay i'll do it right now.. the computer did not boot the last time, let me try again..

[apache_1]

no luck booting up! i tried like 2-3 times. John, i have to go to a sales meeting right now so i think this will be left here again.. i will look for compaq/hp partners around me and try to find out the pricing etc. and next time one word "WARRANTY"!!

[JohnWRuffo]

You can use this form to find a local provider:
http://hp.infonow.net/bin/findNow?LIST_LIST_KEY=2-10936358720021345068&CLIENT_ID=HP_LOC_USA_SRV

I sent you an e- message too.

[JohnWRuffo]

Whoops!
http://hp.infonow.net/bin/findNow?LIST_LIST_KEY=2-10936359870021345136&CLIENT_ID=HP_LOC_USA_SRV&PAGE=Search.html

[apache_1]

Hello John,

I am back again.. I talked to a support tech and I have the numbers for hp resellers close to where I am.. For one last shot I will try something he suggested.. According to him..(in that order)

1. run smartstart and erase and install windows 2000 server..
2. Apply windows service pack 4 (which I had not done initially)
3. enable SNMP
4. Apply Proliant support pack (that you gave me the link for)

The only pieces that I did not try last time were the SNMP and service pack 4.. I will try it with that and see if thats whats making it unstable... If not I'll talk to my local guys and ask them to come over and take a look..

I'll get on it and keep you posted at how the attempt goes.. By the way, i am back at 256 ram (only one dimm) with windows 2000 server installed on it...

[apache_1]

Hi John,

Here is the current situation..

I did all the steps that I mentioned in my last post..

now i have, windows 2000 server + service pack 4 + snmp enabled + psp applied + i ran the dignostic from the diskette. It says no problems...

I am letting it stay running for a while and see what tricks it plays on me this time or do you think that all these service packs and support packs had the elements that make it more stable.. if thats the case, its great.. i should be okay now!!

any advice John??

[JohnWRuffo]

Well... your original errors were certianly hardware based. You could not even boot up.

Definately let it "burn in" over 24 hours before assigning it an application or a task you will trust it to perform.

SP4 is a more mature code to run and the SNMP package will allow the health agents from the PSP to log any errors and such.

You are definately your companies Systems Admin now!

Let me know if I can be of further help? -john

[apache_1]

Hi John,

I will post back tommorow morning with the health report of my box.. Lets hope it stays awake and plays a good boy..

apache (system admin :-)

[apache_1]

Same thing happened!! it was down this morning and can's boot up.. same old same old :-(

i have to go out today but will come back this afternoon and try to do stuff on it again..