Serial Relay Agent keeps losing connection with UPS on COM1

[Marcus Martin]

Hey Glen I did finally get this working on all ups's.

Ensure you are using the excact serial lead that came with the ups, as not all serial leads work :)

I'm reasonalbly sure though that this is not the only issue with ups communications, as had to try several different servers before i found one it would work on.

[glenn calverly]

Thanks. I noticed that too. There are different cables depending on the time of manufacture too. :)

Here's a scenario: I just racked and built 9 new servers in a new rack. 7 380's, and 2 580's. Three UPS are serially connected to 3 of the 380's (with te cables they came with), all build using the same scripts, and Windows 2003, all updates, all updated HP software too. 2 of them failed with the error of:

"The description for Event ID ( 57 ) in Source ( HP Serial Relay Agent ) cannot be found. The local computer may not have the necessary registry information or message DLL files to display messages from a remote computer. You may be able to use the /AUXSOURCE= flag to retrieve this description; see Help and Support for details. The following information is part of the event: ."

. . .in the event logs, minutes after SRA was installed. (nothing follows the error). The error repeats about every 13 seconds. Re-starting the service or re-installing it made no difference.

I've even thought it was the default log size, as some servers this had failed on have the logs set to something other than Microsoft's default.

It's getting out of hand!

[glenn calverly]

Updated, I moved the two that failed to other servers in the racks. One worked. The other failed. I moved that one to yet another server, and it failed with the same symptoms. So, out of 6 servers, 2 work and 4 have failed.

All 3 UPS and cables were brand new.

[Marcus Martin]

Glen I feel your fustration from here, I cant believe there is not a simple solution which would appear to be a not so common but equally fustrating non systematic error.

There is 5500XR PnP driver on the hp site im sure you've found it already but i'm reasonably sure after all the fault finding we have done between us it has something to do with the way the systems are not picking up the UPS's the way their supposed to. More investigation need on this though.

Have Fun

[glenn calverly]

Thanks Marcus. I had found that driver, but dismissed it as incompatible. I'll give it a shot and let everyone know the results.

[glenn calverly]

Ok, here's what I got, and it's interesting. The R5000XR driver did nothing for me. Since these UPS's are basically re-badged Powerware ones, I even thought to look around their website for answers, but to no avail.

But, On my journey, I found an update for the BIOS on the UPS. It updates to version 2.00. So, on the above 2 good UPS's I updated the BIOS without problems. On the bad one, it would not see the UPS, even if I swapped the cable with a known good one. So, I went back to one of them many UPS's that have failed with the same event log error, and tried to update it.

This particular one is a ML570G3, and was installed in November, and failed December 06. I restarted the service, and it came back and worked until Dec 22 before it failed again. So, I managed to update the BIOS successfuly, and the service restarted, successfully!

So, we have one error, which happens for different reasons. It happens on different servers, different hardware, different O/S's, with different serial cables. One problem involved only one UPS as the commonality. The other involved the same error, but with different symptoms.

[glenn calverly]

Sorry, forgot to post a link to the BIOS update . .

http://h18023.www1.hp.com/support/files/server/us/download/25570.html

[Marcus Martin]

Goodness me what a ridiculous state of affairs. So a Firmware update is needed for the UPS's?

I still have a strong tendancy to think towards the PnP driver is incomplete or a nonstable effort on hp's part and maybe they should re-write it to suit the systems they say it supports and include entries for changes in system updates.

The reason i say that is that i managed finally to get all ups's connected without the Firware update although it was a VERY long winded effort.

Well good luck in your Quest Glen, Have Fun :)
Marcus

[glenn calverly]

Hi Marcus,

No, a BIOS update isn't needed, per se, but I wanted to know if it helped. I upgraded these 3 new UPS's (or tried) just to see if it made a difference. It actually knocked one out. I had no problems until after the upgrade, now one of them refuses to communicate on any level, where yesterday it had no troubles. About an hour after the BIOS upgrade, the service started spewing errors in the log, as above.

The BIOS's in my UPS's across the enterprise range from V1.20 to these newest ones at V1.90.

I agree with you about the P&P driver. The reason the R5000XR driver is no good to me is the GPO in this domain forbids unsigned drivers to be installed.

I'm about ready to sacrifice a goat without blemish, if I thought it would fix this problem. ;)

[glenn calverly]

So, the third new server I installed quit today. That makes 7 of 9 brand new servers that won't communicate with the UPS's. There are only 2 left in the rack, so I cannot monitor 1 of the UPS's in this rack.

One of the servers I re-started on the 9th also quit today. It first quit Dec 11.

By some coincidence, both servers quit this morning shortly after 3 am. One at 3:11, one at 3:14.