PE2650 will not boot at all

Hello

The system completes POST before it attempts to initialize firmware from devices like the ERA, so it sounds like it is completing POST but hanging on loading firmware for some device.

I would suggest removing any unnecessary devices like the PERC and ERA. If the PERC or ERA are faulty then it could hang the system. You will need to resolve the hanging issue before you can boot from anything like the floppy.

Thanks

Thanks Daniel,

The ERA is "integrated" in the motherboard for this model ... so i can not take it out ... the PERC 3di is also in the motherboard ... but there is a removable chip there that is a part of the raid ... so i can take it out to see  .... but the machine stops before the bios ever gets to initializing the raid volume or the containers ... .

after the vedio comes on ... the first thing it tries to do is "initialize the ERA" ... it fails at that ... says there was an "error initializing the era" ... then tries to go on ... and at the next step (which should be the machine firmware) .... and it fails there also ... it says configuration is invalid ... need to enter set up ... and then it says "embedded server management is not present" ... then it says there is two 2.4 mhz processors ... and the blinking curser starts ....... nothing else ...

i try clicking the f2 to go to set up ... but it never goes there ....

i know from running a similar model of this ... that the bios does not get to initializing the PERC3di unlill a bit later on down the load up sequence .... so taking out the raid chip from the mother board would not be likely to help ... but i will try that anyway in the morning and see if it does the same thing ...

I'm about to leave for the night and I won't be here tomorrow. If it is possible to disable ERA via jumper then try that. I would look up the system to check but I'm a little pressed for time right now.

If there is no way to disable the ERA and removing/disabling all unnecessary devices has no effect then it is likely that the system board or riser will need to be replaced, depending on if the ERA is on the riser or system board.

Thanks

The ERA is just a rj45 plug attached to some circuits on the motherboard ... there are no cables there ....

 Here is exactly what it does

 1.  power on

2.  vedio on ... says  "ati rage sdram bios r0488 pn gr-s...

3.  says "Remote access controller detected" ...

4.  says "initializing RAC" ... this take s a few minutes ... then

5. it says "ERROR:  RAC initialization failure"

IF this is the first boot up attempt after i cleared the server "NVRAM_CLR" jumper ... then it immediately jumps back to the "DELL" screen and starts over ... if it is the second or later restart attempt it does the same as above but continues on to the next point below

6.  says ... ONE OF TWO THINGS ... not sure if some conditions identify which option occurs or if it is ramdom ..

EITHER

..  "checking embedded server management FIRMWARE" ...

OR

... "invalid configuration - enter setup" ....

EITHER WAY it goes above .. ... it then   says ...  "embedded server management is not present"

7. says ... "Two 2400 mhs processors ..."

8.  blinking cursor starts  .... at this same time the light on the FLOPPY drive a comes on and stays on ...

NOTE:  i have tried clickiing f2  as soon as the vedio starts ... also when it mentions invalid config  .... either way it ends up the same .... same messages ... same blinking cursor at the end .... and it never goes to the set up .... in THOSE twocases, the screen does change to "ENTERING SET UP" ... even though it never goes there .... HOWEVER ... after the blinking cursor shows up .. if you push the F2  ... ... the screen does NOT change to "entering set up" ... nothing else happens after blinking curser showes up .... 

thanks again .... im stumped ...

OK, i removed the PERC3di chip, also the perc ram stick, and the perc battery ... reset the server NVRAM jumper ... and tried rebooting it again several times ...

that change did NOT have any impact at all on how the machine acts ... i did revise my list above to be more precise about what exactly the machine is doing ... and that is it ... it acted this same way both before and after the removal of the PERC chip/ram/battery ...

i reviewed the server manuals and do not see anything about being able to put a jumper somewhere and bypass the RAC ... I WOULD LOVE to know how to do that if there is a way to do it only insiders know ... from what i have read that embeded rac has caused a LOT of different problems for folks besides me ... and i sure do not need the rac features ...

please let me know if you think there is any solution ...

thanks again

I reviewed the system configuration and I'm unaware of any method for disabling the ERA aside from removing the chip.

You have tried all of the steps I would have suggested. I think the next step is to replace the system board.

Thanks

Is there a way to ... and can you explain it .. or refer me to some manual that would explain how to .... CLEAR out the BIOS and machine firmware ... and then reapply that ...??

No

I don't know how to do it, but I know that it requires special hardware and software. This is not something the average IT person can do. A computer shop with specialized equipment might be able to do this.

Is there a way to ... and can you explain it .. or refer me to some manual that would explain how to .... CLEAR out the BIOS and machine firmware ... and then reapply that ...??  it seems to me that something has become corrupted with that basic programming ... and replacing it would solve this problem .... the machine was booting fine ... and then it stopped ... im pretty sure something has gotten corrupted as compared to the motherboard shorting out ....

replacing the motherboard in a box this old is not a viable option ... and used boards are more likely to be genueinely problematic than this one is ....

i appreciate any reference you can give me to do this ...

As Daniel said, wiping the BIOS at the BIOS microcode level or replacing it is not an option.  You CAN however clear the NVRAM, which is the non-volatile memory that holds the BIOS configuration information (not the BIOS microcode itself, but the non-default settings and hardware information).  Sometimes that can become corrupted and cause various issues - clearing the NVRAM clears this information, returning it to a fresh state with default BIOS values.  You do this using the jumpers on the motherboard.  Move the jumper to the NVRAM_CLR pins, power on until you see the message about the jumpers being installed, power off, replace jumper, power back on and see if it is resolved.  If that doesn't do it, then the problem is likely faulty hardware (and/or faulty - and inaccessible - firmware code), and I agree that nothing short of replacing the motherboard (as low as $24 on eBay) will fix it.

Thanks flash ... i have seen those on ebay for $25-30 too ... the problem is ive also read about a lot of people having troubles with this integrated rac (aka ERA) ... and so i wonder how many you might need to try of those to get one that has not also got problems with this same era circuit ... maybe that is why those are so cheap there is a lot of them do not work?

anyhow ... im doing this for an education to start with  ... (you can stock a library with stuff i do not know about these machines) so im going to try one of them and I will post here later on about how it worked out ...probably a week or two to get that done ...

i just assumed (out of ignorance)  there was probably some secret way to wipe the bios and firmware slate clean and start over ...

thanks all for the input ...

At one point, I managed close to 300 2650's and I have only seen this issue 2-3 times.  I have since sold most of those (most locally through classifieds, but many on eBay too) and have never had a buyer mention it (30-90 days of support/warranty on any I sell).  Despite what you may have seen/read, it is NOT a common issue, so chances of ordering one with the same issue is EXTREMELY unlikely (I would say 'impossible', but I don't really believe in that word).  Many eBay sellers will warranty against DOA/defects anyway.

"i just assumed (out of ignorance)  there was probably some secret way to wipe the bios and firmware slate clean and start over ..."

Some high-end gaming boards have BIOS recovery built-in, but that is not "standard" even on modern desktops/laptops, let alone 12 year old servers :)

The closest you will get is the NVRAM CLR jumper.  It would be silly not to try this.  For informational purposes - in some similar cases, you can re-flash the BIOS to fix some issues, but if you can't get the system to complete POST, then that won't work.

Flash,

I'm hoping you can help us.  We have 2 Poweredge 2650 and they've been great for 14 years.  One is now given us boot trouble and it has vital data on it that apparently wasn't getting backed up.  It starts to boot to "Initializing Remote Access Controller", but eventually returns "Error: Remote Access Controller Initialization Failure".  It then continues to load with "Checking embedded server management firmware" and dies at "Embedded Server Management Error". 

Reading other solutions, I've tried reseating the Riser board, but to no avail.  Do any solutions pop into your head?

Thanks,

AJ

The 2nd server that we have is actually just for backup.  It was purchased brand new at the exact same time as the main server.  Could I simply remove the riser board from the 2nd server and install it in the main server, and be good to go?  If so, will doing that lose any of the existing data, or reset any of the setting on the main server?

I greatly appreciate your help.

AJ

You could try clearing the NVRAM using the jumpers on the motherboard. If not, then replacing the riser is about all you can do.

That did the trick.  Putting the other Riser board in the main server got everything back running perfectly with no data loss.  Thank the Lord.  Now I just have to order a new Riser board to replace in the 2nd server.  Certainly good advice.  These things have been workhorses, but likely time to replace them.  I greatly appreciate your help.

AJ