L2000 Memory

[Jim Mallett]

I have an L2000 that currently has 4G of memory (512MB HP memory in 0a/0b thru 3a/3b). To get us by till next year I tried installing Kingston 512MB memory chips(A5798A equivalent) in slots 4a/4b thru 7a/7b.

I think I followed all the rules. Everything was done in pairs, the HP memory stayed in the low slots (in case I had to pull the Kingston memory), all are 512MB so no size issues. Upon reboot the system went into a panic and the GSP reported many errors, mostly along the lines of:

SOURCE: 7 = memory
SOURCE DETAIL: 6 = physical memory SOURCE ID: 0
PROBLEM DETAIL: B = missing

SOURCE: 7 = memory
SOURCE DETAIL: 6 = physical memory SOURCE ID: FF
PROBLEM DETAIL: A = ECC error

SOURCE: 7 = memory
SOURCE DETAIL: 4 = SIMM or DIMM SOURCE ID: 0
PROBLEM DETAIL: D = control address parity error

PROBLEM DETAIL: 2 = single bit error

(full versions attached)

So I thought I maybe had a bad chip. I tried two Kingston in 4a/4b and had the same issue, tried a different two Kingston in 4a/4b, and so on. I was going to pull the HP memory and try all Kingston but my downtime window is just about up. I could see one, maybe two bad chips, but I find it hard to believe all 8 were DOA.

Am I missing anything? From forum reading I see that people have gotten it to work.

Jim

[Luk Vandenbussche]

Hi Jim,

I think you have a physical memory problem.
Are all the modules wel installed in the slots.
The combination of HP and Kingston memory is possible.

You only receive the following error during boot

One or more on HP DIMS were detected.

[Mark Nieuwboer]

Hi Jim,

You have a memory model problem which one it is i can't see you have too look in your log file.
If you have a remote console start the logging. and dump the gsp log. In the gsp/mcp log you can see wich module is going wrong on which slot.

grtz. Mark

[Sameer_Nirmal]

Hi Jim,

What's the PN on the SDRAMs you are using?

What's the PDC and GSP firmware versions?

Are these error are coming for a specific slot ?
What you see in
GSP> CL
GSP> SL > E
BCH> IN > IO
BCH> ME
after installing the new memory?

There is no attachment of the complete GSP log.

I would try with a pair in a slot at a time This is just to verify if the slots are OK. It may be a case where you are trying pairs of new memory in slots 4A/4B and one of them is not OK.

[Jim Mallett]

I didn't even think this post went thru, when I submitted it said "System Unavailable". I ran out of time so I didn't get to try a few things I would have like to, like trying only slots 5a/5b 6a/6b etc.

The attachement is here this time, although it isn't the whole log as I have no clean way of dumping it. At the end of the log is also a cstm output of my working configuration (without Kingston memory).

I've never had the honor of messing with the BCH, so I don't have any information from it. I will do some reading on that before my next downtime window.

GSP Info: Hardware Revision 8 Firmware Revision A.01.10 May 24 2001,11:05:55 (I'll be upgrading to A.01.12 which I think is the best I can do for this)

System firmware = 44.28 (which I think is the latest)

It appears TG Manikandan had the same problem (same error messages) in: http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=141530 but there was no resolution posted.

The info on the memory is:
Kingston KTH-LSVR/1024 (2x512MB)
PN: 9962326-001B00

So I already had 8x512MB HP chips in 0a/0b thru 3a/3b, and wanted to add 8 Kinston chips to 4a/4b thru 7a/7b.

Jim

[Ted Buis]

I presume you are sure it is an L2000 and not an L1000. On the L1000 not all the memory ports are active. I presume you have double checked with Kingston, that the memory you got is supposed to work in the L2000.

[Mark Nieuwboer]

Jim,

For the logging. If you use putty or something like that you can tur on logging.
After that logon the gsp and read the log files. When you read the logfiles al the output is stored in the putty.log file and then you can publish that.
Other question your system is now running without the kingston memory?

grtz. Mark

[timmy b.]

Jim,
If you can get past the alert messages and get to BCH (Boot Control Handler, or "Press any key within 10 seconds to interrupt the boot process"), type in the following command without the quotes, "in me" which is INformation MEmory. This will display your memory and show the locations of any problem DIMMS. Hope it helps!

Tim

[Sameer_Nirmal]

Jim,

As per the GSP log, it could be make out as follows
Memory module 6b ( module which was installed at that time ) has "control address parity error".
Memory modules 0a,0b has "ECC error".

Now the log seems to be conflicting because 0a,0b modules (old) are working OK without new memory right? Did you see such "ECC error" earlier for those modules?
If not, I would take out that 6b module and
see what happens. I maybe wrong , but the memory interleaving causing this??

Moreover,I would go through the complete GSP log (from the fist entry to most recent entry)to have more assessment.

As I said before, it is better to try with a pair of modules and a slot pair at a time.
Then looking for GSP logs for errors if any.