Windows Server 2000, Unknown Prompt with no Code

[Joshuacurtis1]

So, the server at where I work has gone down yet again, there was a storm and the power went out, the server is surge protected on a UPS, so there shouldn't be any damage due to the outage itself. It fails to boot into server 2000. I believe that it is setup in a RAID 1 (not sure if hardware or software RAID), and that one of the drives has failed. If I press Crtl+C on the screen that essentially tells me that the boot has failed (repeatedly attempts to network boot) I am presented with this befuddling prompt, there are no codes or anything else, just this: If anyone has any thoughts on what I should be doing in general, or what this prompt means I would really appreciate it. I am unfamiliar with Server 2000 (I'm only 17 so... not sure how I would be but I digress) so responses with detail for someone well versed in general computer/tech terms, but not well acquainted with older server operation would be appreciated. 

 

I actually have the windows server 2000 CD, is this prompt telling me to insert that? If so, then does that mean I need to repair the windows server OS? If so, will that effectively "reimage" the server, leaving me without my data? I can always take it in to a recovery service or call someone out to look at it but would prefer not to if at all possible. 

 

If it helps, the server is an IBM 7947

The model is AC1

Product ID is 7947E1U

 

 

[dalekphalm]

You must be young.

 

That's a Floppy Disk - the server is basically saying "no boot device found", and is asking you to insert a boot drive (via Floppy, but you can no doubt use any boot device, including CD/DVD - possibly USB, if a server that old even supports USB Boot).

 

If a drive has failed, you likely need to rebuild the array.

 

When the computer boots, does it flash on the screen a RAID configuration option? Normally you press Ctrl + C to open it.

 

You need to replace the bad drive, then start the rebuild.

 

If the array cannot be recovered, you'll need to replace any bad drives, create a new RAID array, and reinstall from scratch.

 

Hopefully you've been backing up any important data...

[Joshuacurtis1]

I am indeed young, I said so in my original post. I'm aware the image is of a floppy disk, the server has no floppy drive, which is why I wasn't sure if I was supposed to insert a new drive or boot media.

 

Just so I understand you, you're saying that this prompt is requesting I insert the installation media (in this case a CD) and allow the server to repair itself, and there is then a chance that the drive will repair itself, and not overwrite the data on the array? 

 

Yes, a prompt does appear on startup asking about the RAID configuration.

 

We have been doing backups, but would like to get the server online nonetheless.

[dalekphalm]

4 minutes ago, Joshuacurtis1 said:

I am indeed young, I said so in my original post. I'm aware the image is of a floppy disk, the server has no floppy drive, which is why I wasn't sure if I was supposed to insert a new drive or boot media.

 

Just so I understand you, you're saying that this prompt is requesting I insert the installation media (in this case a CD) and allow the server to repair itself, and there is then a chance that the drive will repair itself, and not overwrite the data on the array? 

 

Yes, a prompt does appear on startup asking about the RAID configuration.

 

We have been doing backups, but would like to get the server online nonetheless.

The image is prompting you to insert a boot device. This could be the installation media, or it could be a HDD with Windows Server 2000 already working. Basically, that image just means "no boot device found".

 

Second: Repairing the RAID and using the Windows 2000 installation Media are two totally different and unrelated things (Depending on how RAID was setup).

 

If RAID was setup using built-in motherboard RAID, OR through a Hardware RAID card, you'll need to repair the RAID array BEFORE doing anything with the Installation Media.

 

I do not believe that Windows can create a bootable RAID1 Mirror that has the OS on it, so odds are, the RAID Array was created using Motherboard BIOS, or a hardware RAID card.

 

You'll need to figure out which first before doing anything with the installation media. Because you see a prompt about RAID configuration, I'm leaning towards a Hardware RAID Card.

 

Enter the RAID Configuration page, and look at the array status (Probably "degraded"). Find out which disk dropped out of the array - replace with a comparable disk (same model ideally, if not, same RPM and same capacity). Once replaced, start a rebuild.

 

After the rebuild is complete, restart the server and see if it just boots up normally.

[Joshuacurtis1]

Ok, now I understand what that screen means, thank you. 

 

I know they are unrelated, I didn't configure this server (obviously) so all I was told is that the server has 2 drives in it, and that "They mirror each other." This led me to believe that they were in RAID 1, however I am not certain. 

 

I have 3 questions then:

 

How would I check to see if either of the drives have in fact failed, and if yes, which one, or both? 

 

How would I know if the server is configured in RAID, or, if its even possible, that the server is using something else entirely? 

 

How do I actually rebuild the array? (If you know of a good guide or other post to read, I can just follow the instructions from there)

 

 

[dalekphalm]

4 minutes ago, Joshuacurtis1 said:

Ok, now I understand what that screen means, thank you. 

 

I know they are unrelated, I didn't configure this server (obviously) so all I was told is that the server has 2 drives in it, and that "They mirror each other." This led me to believe that they were in RAID 1, however I am not certain. 

 

I have 3 questions then:

 

How would I check to see if either of the drives have in fact failed, and if yes, which one, or both? 

 

How would I know if the server is configured in RAID, or, if its even possible, that the server is using something else entirely? 

 

How do I actually rebuild the array? (If you know of a good guide or other post to read, I can just follow the instructions from there)

 

 

1. To check that a HDD has actually failed, you'll need to run a HDD testing software on it. Each HDD manufacturer has their own tool for that - eg: SeaGate SeaTools, WD Data Lifeguard, etc. So find out what HDD make it is first. You can plug the drives into another computer, load up the software, and run a long self test (non-destructive). It'll check for bad sectors, read errors, etc.

 

2. You need to check that RAID Configuration prompt you saw earlier. It should clearly list any Arrays created (working or degraded). Take some pictures and post here if you're not sure what you're looking at.

 

3. The rebuild process entirely depends on how the RAID array was created, and what brand (Eg: LSI vs Adaptec, etc). If you can answer question 2 first, that'll help you figure out question 3. Generally, you select the array, and there should be a "rebuild" option or button.

[Joshuacurtis1]

1 The drives use a SAS interface, not SATA and this server is the only computer that I have access to that uses a SAS connector, so that may take a while.

 

2 I was incorrect before in saying that there was a screen that prompted me for any RAID setup, but in IBM's diagnostic software for the server which I did manage to access said that there was an LSI SAS controller, assuming this means that is a RAID card

 

3 In the event that I figure out how to access that, I will try to rebuild the array. 

 

On startup the only prompt I am presented with is the attached image, would the RAID setup be in more of a command line text format, as opposed to one with graphics? 

[Joshuacurtis1]

11 minutes ago, dalekphalm said:

1. To check that a HDD has actually failed, you'll need to run a HDD testing software on it. Each HDD manufacturer has their own tool for that - eg: SeaGate SeaTools, WD Data Lifeguard, etc. So find out what HDD make it is first. You can plug the drives into another computer, load up the software, and run a long self test (non-destructive). It'll check for bad sectors, read errors, etc.

 

2. You need to check that RAID Configuration prompt you saw earlier. It should clearly list any Arrays created (working or degraded). Take some pictures and post here if you're not sure what you're looking at.

 

3. The rebuild process entirely depends on how the RAID array was created, and what brand (Eg: LSI vs Adaptec, etc). If you can answer question 2 first, that'll help you figure out question 3. Generally, you select the array, and there should be a "rebuild" option or button.

Update: I removed the drives from the server, then reinserted them and it booted correctly, with all volumes and data shown correctly... Bizzare but I'm not going to question it. Thank you for all your help, and the information you provided, it will allow me to do much more in the future!

[dalekphalm]

1 minute ago, Joshuacurtis1 said:

Update: I removed the drives from the server, then reinserted them and it booted correctly, with all volumes and data shown correctly... Bizzare but I'm not going to question it. Thank you for all your help, and the information you provided, it will allow me to do much more in the future!

Sounds like one of the drives simply dropped out of the Array.

 

I would still try to test both of the drives - one of them might be dying.

[Joshuacurtis1]

Just now, dalekphalm said:

Sounds like one of the drives simply dropped out of the Array.

 

I would still try to test both of the drives - one of them might be dying.

Interesting, I wasn't aware that that was possible

 

I certainly will, going to do a complete backup right now just in case, but that's next on the list.

[dalekphalm]

Just now, Joshuacurtis1 said:

Interesting, I wasn't aware that that was possible

 

I certainly will, going to do a complete backup right now just in case, but that's next on the list.

Sometimes a perfectly functional disk can get dropped out of an array if, for example, it takes too long to respond to the RAID controller, or encounters some other hiccup.

 

It's also possible the RAID controller tried to write to a bad sector. A few bad sectors are normal for an old drive - HDD's have spare sectors that are used to remap the bad ones, but there's only a small number of spare sectors. Once the spares get used up, the drive is basically dying.

 

Any bad sectors generally point to a drive that's aging, and should be replaced soon though - personally, I'll replace a drive as soon as any bad sectors are discovered. You can use a drive with bad sectors for non-critical stuff, but don't put anything important on them.