3

Walking by the server closet I noticed that one of our servers (namely the one that hosts the company software/data) is beeping: low-high repeat. As soon as it started, I backed up all of the data stores to separate drives, but it's been fine since it started beeping yesterday afternoon.

Now I'm only the lowly software developer and our server/network admin is n/a so far, and no desktop machine I've ever worked on or with has had any serious problems before.

Details and Reiterated info without my dialogue:

  • Beeping is low-high reapeating
  • Intel Core2 Quad machine (if it helps)
  • It's not one of the UPS'. I made certain that it was indeed the server beeping, and it was obvious considering that it's about 4ft high on the rack and the UPS' are on the floor.
  • Runs Win Server 2k3
  • Again, everything is still working fine. Even when I remote into the machine, there's no errors - nothing.

Not sure how to tag this question, feel free to edit.

4
  • What is the make/model of the server?
    – DanBig
    Aug 19, 2009 at 15:31
  • 3
    LOVE the answers to this question. Between everybody I think we've pretty much "guessed" every component except the chasis lid. So I'll go ahead and guess that one just to cover all bases. ;-)
    – KPWINC
    Aug 19, 2009 at 16:10
  • Oh yeah! That's obviously the "tertiary fan disconnected" beep code for the HP 6000SUX! I'd know that beep code anywhere by just the server's position on the rack and the vague description of the beep!
    – Ernie
    Aug 19, 2009 at 19:59
  • 2
    @Ernie... Sorry for being ambiguous, but being a software dev with a boss that's too cheap to hire a sys/network admin, I have to figure these things out on the go and I'm trying my damndest. Oct 13, 2009 at 18:50

10 Answers 10

4

RAID array failure? Second power supply failure?

3
  • RAID array failures are usually a constant tone, at least, from what I've heard.
    – DanBig
    Aug 19, 2009 at 15:49
  • 1
    It's completely hardware dependent, and there are no standards. My experience, FWIW, is the opposite.
    – tomfanning
    Aug 19, 2009 at 16:52
  • Yea, one of the drives in the raid failed. I was able to swap in another and rebuild the array. Thanks for the tip Oct 10, 2009 at 6:59
9

Well, here's what I would do:

  1. Find the manual for that make and model of server, or for its motherboard. They're often available on the supplier's website if you can't find the dead tree version.
  2. Look up the definition of the beep code in there. Failing that:
  3. Call the vendor's tech support line.

Any beep codes are highly hardware dependent. Even PC BIOS beep codes are non-standard.

5
  • +1 for the 'heres what I would do' approach
    – cop1152
    Aug 19, 2009 at 16:14
  • +1 for the dead tree joke.
    – IceMage
    Aug 19, 2009 at 16:16
  • This works most of the time. google: {manufacturer} "beep code" Where {manufacturer} is the company that provided the box e.g. Dell or the motherboard or whatever part is beeping (i.e. raid controller).
    – Chris Nava
    Aug 19, 2009 at 17:02
  • Dead tree! thats pretty good.
    – Alan
    Aug 19, 2009 at 19:26
  • turns out one of the drives in the raid failed. Replaced it, rebuilt it. Problem solved. Thanks for the help. Oct 10, 2009 at 6:58
4

Beeping is normally an indication that something needs your attention, but - as you've seen - determining what that something is can be a whole other world. You need the make and model of the server first and foremost, as the beep codes could be different depending on these factors.

Now looking at the logic of it, the server is still running, so whatever it is that needs your attention is something that the server can remaining operating despite a failure or imminent failure of. This indicates something that there is more than one of, commonly either a hard disk, a CPU (if more than one), a memory bank or a PSU.

If it's a HD, and again depending on the make and model, there may be lights on the front panel to indicate this, and which HD it is. You'll be looking at a front-mounted drive caddy, with each HD in it's own slot, and each with at least one light, which may be green, orange or red.

If it's a PSU these will slot into the back, and should again have lights. You can recognise these cos the power leads are going into them. ;)

If it's one of the others, there may or may not be a front status panel giving more info, or even just an error code (e.g. something like e1211 if it's a Dell). This may also have either a green, orange or red light or be backlit in one of these colours.

There may also be a diagnostics program installed on the server which could pull out the info, but to run that you'll probably need Administrator access which I'd guess you don't have.

So based on this you can do a certain amount of fault diagnosis, at least to the extent of determining what it's not. After that - STEP AWAY FROM THE SERVER. Kudos to you for trying to be helpful, but if you try to do any more work here you'll be a person who is not authorised to do this work who is taking on responsibility for something that ain't your area of expertise, so you need to hunt down your sysadmin and explain why he owes you a beer. :)

1

Does the Event Viewer show anything? Many brands of servers have utilities installed from the manufacturer that can show where the problem could be.

1

It's definitely a failure of some sort. Can be a variety of things, but definitely worth having the Admin look at ASAP.

Possible problems could be: RAID Disk Failure, ECC Memory Error, Fan Failure, Power Supply Failure, Case detected as open, system is running too hot. There are many more, but as you can see, there are many many things that could be wrong. You'll want to make sure the admin knows about it as soon as he becomes available, and may want to notify him immediately.

Also, remoting into the machine will likely not show you the errors, since they are probably hardware based. You'll be able to see them if it's a dell and you run the dell system utility (windows based), or a server specific management program that lets you view the status of a RAID array, fans, system logs, and the such.

2
  • To expand one of your answers based on my first instinct - it could be a redundent power supply failure (one of the 2 power supplies has failed). RAID error is another but seems less likely because OP has not seen any secondary error notifications (lights, event messages). Case open (ajar) seems to be a pretty likely too.
    – David
    Aug 19, 2009 at 15:48
  • If it's a power supply failure, you don't have much to worry about since power supply failure is fairly rare (more rare than disk failure at least) and a power supply that is running solo is actually more power efficient. A case open problem would be more concerning to me since that can disrupt a rack server's internal air flow, causing it to overheat (bad).
    – IceMage
    Aug 19, 2009 at 15:57
1

Yes, notify the admin - this is their turf. If you still want to help after that then cool, but CYA.

Find out the model number, serial number and make of the server(Dell 2650 Edgeline).

Open msinfo32.exe and look around for the motherboard information it'll probably simliar to: I/O Port 0x00002000-0x00002FFF Mobile Intel(R) 45 Express Chipset Series PCI Express Root Port - 2A41

Baring that, open the server up while on(not best practice) to locate exactly which part is beeping(I'm 95% sure it's the mobo though).

Then look up error codes for the beeping part on the manufacture's website. The model number, serial number and make will now become useful.

Don't expect it to be exactly obvious but the erorr codes will give you a pretty good idea of what's wrong.

From there find out if it's under warnety and call tech support.

But again, this is all the admin's truf so CYA and tell him about it.

1

Not the answer you want but I'd call the vendor responsible for the hardware/warranty - it should be on an "easily" readable sticker on the server itself ;) (sometimes people put them on top of rack servers and then push another server into that slot, hiding the sticker ><)

It could be anything, you need to at least find out the make and model in question....

1
  • +1 for having those we pay to do these things do them. Aug 20, 2009 at 2:39
0

This could be a multitude of failed things. Memory, CPU bank, PSU, Hard Driveor FAN. My guess is on the PSU or fan. The options are these:

  1. Get the Serial number for the server, find the manual online and see what the beeping could indicate.

  2. Look to see if there is a LED panel. Usually beepings are with LED alerts. Or any Hard Drives have Lights.

After saying all this, and realizing that IceMage said most of this as well... know that if the server is still working, it probably is just in a degraded state. Open a ticket or Email your server support team so it is documented, then personally talk to the Server Admin.

0

Low-High beeps are commonly a CPU or other thermal sensor trigger alert (ie, overheating). I have also seen raid controllers beep low-high on a failed cache battery, or a failed drive condition.

0

Could this be caused by a failing cmos battery on the mobo? I dont know if the status of this type of battery is even monitored or not, but I havent seen this mentioned. I think they have an expected life of two years. I have never heard a comptuer emit this exact sound.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .