Proliant DL160 G5 Issues.

Vince

Vince

Vince

Man of Honour
Joined
30 Oct 2003
Posts
13,801
Location
Essex
Hi All,

I have a few issues with the above server, firstly since purchasing the server it has been gaining time, in a month it will clock up +10 mins over the actual time. Because we are using one of these as a domain controller the problem is noticeable.

So today I called HP about this who advised that an updated bios should do the trick, reluctantly I agreed and ran the update, now I have a far worse issues. The fans on the server are stuck at 100% and can be heard all around the building.

I called HP who have been, up until now, about as much help as a chocolate tea pot. Does anybody out there have one of these, has anybody experienced these issues, and more importantly does anybody have a fix.

I have tried rolling back to the original bios and have tried every bios right up to the most recent without any success. I have also tried adjusting the fans with command line using the following:

ipmitool.exe -I ms bmc reset warm

This seems to work with some of our other servers and tends to bring the fans down to 3200ish rpm rather than around 8k+ that these sound like they are sitting on at the moment.

OS is server 2008 x64 with Hyper-V any help would be much appreciated.

Cheers guys.
 
Last edited:
Hi there,

Thanks for the reply. I have actually tried the method that they describe in that thread which is basically what that command line is about in the first post.

To reply to the other questions:

Is your CPU running at 100% constantly/A lot? - Very rarely is the server running at anything over about 25%

Are you running VM's on it? - No, No VM's on this server.

Is the DC you describe a VM? - No.

I’m pretty much stuck on this one; the server is simply used for authentication for 100(ish) users. This includes authentication to our offsite Sharepoint farm and also authentication to our offsite accounting system.

As it looks right now, I have been waiting on the phone for HP support for 40 minutes probably to get fobbed off with some other "fix" for the poor bios updates.

As it stands I think that I am going to demand a replacement main board as on boot the server also seems to have inherited another issue:

"611 System Event"

Overall I am pretty unhappy, once the new board is in I think I will go the easy route and point the server at an internet time source as you mentioned.
 
A 611 error points to the system log being full, I'd definitely recommend checking the system log, it sounds like there is something alerting persistently that needs afixin'.

If you have installed the server using SmartStart (Not sure if it's available on the 100 series) then use the system management homepage:
https://servername:2381

If you have an ILO card in the server, then I believe access via that will give you access to the Integrated management Log.

One way or another you need to check those logs, it may even give you some ammo to use on the phone with them.

My final recommendation, is to use HP's Business support portal to log calls.
You can track existing calls and any progress, and update it easily with any information, without having to get through their telephone system every time.
 
Right looks like HP have finally caved in and are sending an engineer who will apparantly be here first thing in the morning with a new motherboard and new lights out controller (Which is commonly the cause of this issue).

However even though we have many HP servers I am really dissapointed with the support, its not an issue with me that the call centers are now India based, but what is a problem is the line quality which I have to say was poor at best, this resulted in poor communication and that constant "Could you please repeat that" backwards and forwards.

Ok granted they are providing the part and engineer but in total today I have been waiting in phone queues for more than 3 hours. To add the mandatory injury to insult I now also have a head ach from my loud arse DL 160.

Ill let you know where we go from here!
 
I have the same problem with HP telephone support, which is why I've now taken to logging calls through their portal.

As an example, we had a yellow warning light on the front of the server.
I checked the system management homepage which said the memory was degraded, stuck the call on the portal, it got approved for parts immediately, and dispatched.

I know this is a more clear-cut case of a Hardware fault, but not having the agony of the language barrier, poor line quality, the wait through the telephone system, and most of all being able to track all the progress of the call online was a massive improvement.
 
Next time I will be taking that route thats for sure!

With the 100 series it looks like we need to power down again and enter the bios in order to look at and clear out the system event logs. This is something I am not going to bother doing, whats the point when new hardware is going to be here at 10:30 am?

Before the bios update however there was no sign of this problem or any issues on post, could there have been a problem during the update that is filling the logs? - I don't know and im sure as hell not spending any more time finding out.

What amazes me is the issues caused by an update of the bios, you would have thought that these things are fairly well tested. I can only say that the new bios will not be tuched regardless of a small niglle such as keeping time.
 
I'm not sure how big the logs can grow to.
It's not necessarily anything to do with the update, it could have been an ongoing Hardware issue before that, but the log has only just now filled up.
For instance if it has 2 power supplies, but only 1 is powered, you'll get an event saying that power has failed to PSU2, or if there is a correctable memory issue, that would be logged.

It could be filled with errors relating to your time issue, it may be utterly unrelated, the only way to know is to check (which the HP engineer may very well do).

Being unable to check the system log online without shutting down the server seems crazy, but I guess the 100 Series are an entry level system and all possible costs have been reduced.
 
Sounds like that could well be the case.

To be honest the server is well over spec for what it is/has been doing already, it never breaks a sweat with its current roles and never really did while it was being used as a file server during our sharepoint implimentation.

I must admit up until the bios update its been an absolute rock and there has not been a single issue (appart from the time). I put some of this down to server 2008 which is, imho, strides ahead of 2003 especially when it comes to disk management.

Anyway Little_Crow thanks for your help, I think that I am going to head off home and watch the footie.

I will send an update tomorrow to let you know how we get on as well as update on what was going on with those logs.
 
Thought I would update quickly, got woken up at silly o’clock this morning by a concerned colleague telling me that some bad noises were coming from our server room. The fans seem to have got louder and louder over night they said (that or less people in the building so they seem louder).

Anyway I dragged myself in expecting to arrive to a server engineer and a new motherboard. The motherboard arrived at 9:30, but no engineer, not a problem I thought the engineer wasn't booked until 10:30.

10:30 came and went as did 11:30, 12:30 and 1, untold amounts of phone calls and still no engineer.

We were meant to be hosting an event today at 2 so things are starting to get tense. You can imagine the chairman and COO leaning on me with the old “you promised this would be fixed” - there is me on the phone to HP trying to fix this mess.

It gets to 1:15 and still no engineer, so I finally ring them to find out the warranty info when replacing the parts myself. After 20 minutes I finally get a response - they tell me that they will honour the warranty in this instance.

15 minutes go by and finally 10 minutes before the event starts there is a silence across the office, as I press the button the server comes back online with a whisper of noise to tell me its booting. A quick change of a few settings in our routers and we hit the ground running (phew).

So what have I learned from this? Well a few things actually, firstly HP couldn’t arrange a **** up in a brewery (especially when you call them on the phone), secondly, if you want a job doing properly - do it yourself.

Anyway the overall HP customer service gets 2 out of 10 from me and that’s only because they managed to get the motherboard out on time. Other than that it’s more of the same, another good company being let down with empty promises and poor customer service. Poor show!
 
I had a DL380 do the exact same thing with the fans 100% when I installed win2k8 phoned em up and a helpful chap had me reset some pins on the board which in turn killed the mobo. They came out the next day and replaced it mind but it took the engineers some time. Worked like a charm after that. Strange problems with HP eh.
 
Strange stuff indeed, I think there seems to be a quality control issue with the new bios releases, I have calmed down a fair bit now and filled in the HP questionnaire as honestly as possible, I can assure you the customer service rating was below par.

On a good note I am quite happy with the new board (even if the engineer failed me).

The fans are now quiet and its running the latest revision of their bios, so far there are no issues with the clock gaining time so thats a plus. Either way it will work out quite costly for HP if every customer that tries to update a bios ends up replacing a board.

I have similar issues on some other HP servers but am pretty reluctant to update any bioses even if it is the recommended step by HP.
 
I just thought I'd give an update on this issue, at least from my side.

We recently bought the DL160 G5 Server and the first thing I did was to update the bios to the latest version. Then all the Fans got stuck on MAX Speed and refused to relax. I phoned HP up and explained what I had done and what the problem was. They recommended I should try to downgrade the bios and then clear the Cmos.

It didn't help. So they sent out a technician to us. The Technician told us to check if the BMI was up to date. So we checked and it wasn't. (The support didn't know about this). The BMI is supposed to be upgraded in the Bios update but for some reason it latest bios had left the BMI in a locked down state and in an old version.

So we downloaded:
Firmware Upgrade for HP ProLiant DL160 G5\DL165 G5\DL185 G5\DL160 G5p\DL165 G6 Lights-Out 100 Remote Management (For USB Key-Media)
Version: 3.11 (B) (29 Jun 2009)
From: Here
Then we applied this update by using a USB-key. Once that was done we updated the bios back again to the latest version.

Believe it or not but that did the trick for us. So if you're reading this post because your server is stuck @ Max fan speed, give this a try.

Good luck :)
 
Good work on the fix Radx... This issue gave me a headach (literally). Mine has been fine since the new board, I am amazed that the support didn't know that this issue existed, it seems like a big waste of money to be sending new boards and engineers out for something that can be fixed fairly easily.

Anyway thnx for the update.
 
You must log in or register to reply here.
Back
Top Bottom