ML110 Restarting Problem!

[Blaz Kuder]

Greetings!

I have the problem with ML110 Server with 4x 1GB memory.
The server restart after few minutes of running.
I tested the Memory, and an error was on address 000000004f8. So I tested each memory separated, and the same error shows up for each memory.
Then I put the memory on the different Memory slot, and the same problem shows up for the each memory.

After that I use the different memory test tool, and the problem stay the same.


So I assume, that the Motherboard is broken, because 4x 1GB RAMs(made for ML110 server) with the same error is 1:1000000, so I order new Motherboard.


After changing the Motherboard, the problem is still here.

Server still restart.

I don't know about memory tests, because didn't test them yet(will do it first thing on the Monday).

So my question is, if anyone know what it could be wrong? CPU? Power Suply?


Will be very glad if someone could help me to solve this puzzle.


Thank you very much!


Have a nice weekend.

[Matti_Kurkela]

Please let me confirm: is the memory error always indicated at the same address?

If so, I'd bet on the CPU being faulty.

Rationale:
- We know that moving the memory around does not move the problem: indicates the problem is probably not in the memory per se, but somewhere on the memory bus (CPU included)

- Changing the motherboard replaces everything on the memory bus except the CPU and the actual memory modules. If this does not help, the problem is somewhere in the remaining components.

- Power supply faults would generally indicate errors at random addresses: the PSU simply does not have anything fast enough to synchronize with the CPU and produce errors that always re-occur at the same address.

- A fault in the CPU's internal cache or in the cache control logic might certainly produce an error that occurs only in some precise address.

MK

[Blaz Kuder]

Greetings!

Yes the error on the memory is the same.

Will order a CPU, and will let you know if this was the right solution :)


Thank you very much!

[Blaz Kuder]

Greetings!

Today I made few more tests.
I changed Memory, CPU, Power Supply, DVD-RW, and the problem is still the same.
When I put Memory, CPU, Power Supply, DVD-RW from problematic Server to one which works, there isn't any problem at all.

Is it possible that the NEW Motherboard which I replaced is in failure?

What should I do?!?


I will order new(3rd) Motherboard, and will made more tests.


Thank you very much!


Have a nice day!

[Blazhev_1]

Hi Blaz,

you didn't mention the Generation, so I assume you have the ML110 G1.
The right memory you must use is : 354563-B21.

Is the server restaring every time while trying to initialize a component(at the same place) or random.
Did you try minimal configuration? Remove all PCI cards, disconnect all external components and the HDDs and try with one CPU and one DIMM. It can be that there is unsupported device.
Since you have already replaced the systemboard and the DIMMs I don't think the problem is in them, maybe unsupported HW or thermal issue(heatsinks or fans), check temperature in RBSU.Are all DIMMs recognized in POST?
If the server restarts how do you test the memory - on a different server?
Do you use Insight diagnostics or you don't have the opportunity?
Do you have other server (the same model) to test the components there? I think minimal configuration will help you find out a solution, since the systemboard, CPU and DIMMs are replaced...

Regards,
Pac

[Jeff Law]

Did you ever get a resolution to this problem?

The reason I ask is that I have an HP ML110 G4 server with 2 x 1GB memory modules, and after working fine for about 6 months, running SBS 2003, it stared locking up.

The hard disk light would be on continuouly, the monitor would switch off saying no signal, and there is no response at all to keyboard or mouse. Wouldnt respond through the network either. Switching the power off was the only option available, where upon it would boot up and continue working until the next time. It got to the stage that it would do this about every 4 hours or so.

I initially thought it was an OS thing as it started happening shortly after an MS update was installed. I have reinstalled the MS SBS 2003 from scratch and it still happened - even before installing the patch that was installed just before it started happening.

I resorted to doing various hardware tests and when I tried using memtest86 v3.3, this would always report an error at the 000000004f8 address. I also went through the exercise of swapping memory, removing memory etc, but memtest86 would always fail during the same test at the same time.

I thought maybe it was memtest that had the problem, and tried it in two other servers, neither of which are HP's, and it quite happily run throught the night with no reported problems at all.

Phoned HP, and I have had the motherboard an memory replaced. memtest still reports an error at the 000000004f8 address.

I have now just had the CPU replaced, and guess what? I still have memtest reporting the same address being faulty.

The guy who replaced the bits doesnt know what else it could be. Neither do I. Hence my interest in whether you have resolved this.

In some ways it is comforting to know that someone else has also had the same problem. It would be more comforting to know what the solution is :-)

Regards
Jeff

[James ~ Happy Dude]

Hello Jeff,

HP will not validate a 3rd party diagnostic tools. Its strange, indeed; that this tool will report an error even after the major components are replaced.

I will suggest you to Download the HP Insight Diagnostics from :
http://h18023.www1.hp.com/support/files/server/us/download/27487.html
& run a test & let us know about the reports.

Regards,
James.

NB: Please restrain from Hijacking other's thread. Please start a new thread for new issues & you may give the links to the existing threads for same issues. Be a Good ITRCizen. :-)