Jump to content


 


Register a free account to unlock additional features at BleepingComputer.com
Welcome to BleepingComputer, a free community where people like yourself come together to discuss and learn how to use their computers. Using the site is easy and fun. As a guest, you can browse and view the various discussions in the forums, but can not create a new topic or reply to an existing one unless you are logged in. Other benefits of registering an account are subscribing to topics and forums, creating a blog, and having no ads shown anywhere on the site.


Click here to Register a free account now! or read our Welcome Guide to learn how to use this site.

Photo

DELL T7400 - alert! uncorrectable memory error


  • Please log in to reply
14 replies to this topic

#1 rotoq

rotoq

  • Members
  • 7 posts
  • OFFLINE
  •  

Posted 13 February 2017 - 03:26 AM

Hi,

 

I recently acquired my friends old T7400, which although it hadn't been used in a few years, was working before it was left in the corner.

 

Running Ubuntu 16.04 LTS, Dual XEON E5420 2.5Ghz, 20GB RAM

 

Symptoms:

 

I am getting this error message on boot:

 

alert! uncorrectable memory error has been previously detected in dimm 1 or 2. Press F1 to continue...

 

The dimm slots referenced also, change from '1 or 2' to '5 or 6' etc. I have swapped all the ram out, checked the contacts, and re-seated it.

 

DIMM1: 4GB

DIMM2: 4GB

DIMM3: 4GB

DIMM4: 4GB

DIMM5: 1GB

DIMM6: 1GB

DIMM7: 1GB

DIMM8: 1GB

 

My PC also hangs, becomes completely unresponsive in every way and has to have a hard reset.

 

I ran memtest86+ for about 3 days, making over 8 complete passes of the 20GB of memory, but didn't find a single error.

 

I don't think its anything wrong with the RAM sticks, as I have tried a few combinations, and the error messages don't repeatedly flag the same 2 sitcks. It's highly unlikely that the 4x1GB sticks that came with it are damaged, as well as the 4x4GB sticks I just bought from a reputable seller who reported them as tested.

 

I have removed and reseated the CPU. I have also replaced the cmos battery, but the problem persists.

 

Any ideas people?

 

I'm going to update the BIOS tonight, in case that's the cause.

 

Many thanks in advance.


Edited by rotoq, 13 February 2017 - 04:21 AM.


BC AdBot (Login to Remove)

 


#2 SleepyDude

SleepyDude

  • Malware Response Team
  • 2,995 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Portugal
  • Local time:01:59 AM

Posted 13 February 2017 - 04:55 AM

Hi :welcome: to BleepingComputer,

 

I did a search but coudn't find information from Dell about the correct order to install memory on that machine...

 

My suggestion put the 1GB DIMM's on slots 1-4, the BIOS shows correctly the information about the installed dimms?

 

If the machine doesn't have the latest available BIOS version I would update.


• Please do not PM me asking for support. Post on the forums instead it will increases the chances of getting help for your problem by one of us.
• Posts in the Malware section that are not replied to within 4 days will be closed. PM me or a moderator to reactivate.
• Please post your final results, good or bad. We like to know! Thank you!

 
Proud graduate of GeekU and member of UNITE
___
Rui

 
 


#3 rotoq

rotoq
  • Topic Starter

  • Members
  • 7 posts
  • OFFLINE
  •  

Posted 13 February 2017 - 05:14 AM

Thanks for the reply. I have tried the 1GB's in 1-4 and vice versa. I've tried excluding the 4x4Gbs, and excluding the 4x1GBs aswell.

 

I'll cross my fingers on the BIOS update for now.


Edited by rotoq, 13 February 2017 - 05:30 AM.


#4 SleepyDude

SleepyDude

  • Malware Response Team
  • 2,995 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Portugal
  • Local time:01:59 AM

Posted 13 February 2017 - 06:45 AM

I'll cross my fingers on the BIOS update for now.

 

Unfortunately the dell information is not very clear about what the update does!


• Please do not PM me asking for support. Post on the forums instead it will increases the chances of getting help for your problem by one of us.
• Posts in the Malware section that are not replied to within 4 days will be closed. PM me or a moderator to reactivate.
• Please post your final results, good or bad. We like to know! Thank you!

 
Proud graduate of GeekU and member of UNITE
___
Rui

 
 


#5 rotoq

rotoq
  • Topic Starter

  • Members
  • 7 posts
  • OFFLINE
  •  

Posted 16 February 2017 - 05:18 AM

Updated to A11 and the same thing happens. Again, changes which DIMMs it references. One time it'll be DIMM 1 or 2, the next 5 or 6 etc. Always in pairs.

 

So I've tried:

  • new bios
  • new cmos battery
  • fresh Ubuntu install
  • reseated cpus
  • RAM tested extensively with memtest86+

Anyone have any more ideas? I can't afford to buy a new motherboard just to test. In any case, it'd be dead money spending £70 on an old motherboard for this, I'd rather save up and build a new server. The point was this being free, cost would be very low.

Frustrating.


Edited by rotoq, 16 February 2017 - 05:24 AM.


#6 rotoq

rotoq
  • Topic Starter

  • Members
  • 7 posts
  • OFFLINE
  •  

Posted 16 February 2017 - 05:27 AM

Hi :welcome: to BleepingComputer,

 

I did a search but coudn't find information from Dell about the correct order to install memory on that machine...

 

My suggestion put the 1GB DIMM's on slots 1-4, the BIOS shows correctly the information about the installed dimms?

 

If the machine doesn't have the latest available BIOS version I would update.

 

Yes, it shows all the information correctly. I believe I've tried the 1GB in 1-4 and the 4GB in 5-8 but I will try again.



#7 SleepyDude

SleepyDude

  • Malware Response Team
  • 2,995 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Portugal
  • Local time:01:59 AM

Posted 16 February 2017 - 05:44 AM

The DIMM's have all the same specs besides the capacity?

 

The BIOS have some option to disable ECC? If possible test with ECC disabled.


• Please do not PM me asking for support. Post on the forums instead it will increases the chances of getting help for your problem by one of us.
• Posts in the Malware section that are not replied to within 4 days will be closed. PM me or a moderator to reactivate.
• Please post your final results, good or bad. We like to know! Thank you!

 
Proud graduate of GeekU and member of UNITE
___
Rui

 
 


#8 rotoq

rotoq
  • Topic Starter

  • Members
  • 7 posts
  • OFFLINE
  •  

Posted 16 February 2017 - 06:55 AM

Running MemTest86 now instead of MemTest86+.

 

Will double check the RAM specs, but not sure it matters as problem occurs with 4x1GB or 4x4GB or with all RAM sticks included. Think there's something up with the memory controller.

 

Going to try and run with each CPU one at a time after that.

 

It isn't possible to disable ECC, thanks for the suggestion though.

 

SIDE QUESTION: How many passes is 'normal' for one to assume RAM is OK? I'm fairly sure the RAM is OK as MemTest86+ made about 8 passes without a single error, and the uncorrectable memory alert is referencing different DIMMS, and it's not very liekly that there is something wrong with ALL of them AND MemTest86+ didn't find a problem with ANY of them.

 

Also: Downloaded ultimate boot CD and mounted it to USB. I'm going to throw every diagnostic I can at this POS until it spits out some answers.


Edited by rotoq, 16 February 2017 - 07:29 AM.


#9 hamluis

hamluis

    Moderator


  • Moderator
  • 55,550 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Killeen, TX
  • Local time:07:59 PM

Posted 16 February 2017 - 09:30 AM

I would remove all RAM...then blow out the memory slots for dust, etc.  Then I would test 2 of same 4GB sticks, usiing Memtest 86+, running one complete pass.If no error message or errors detected...I would go on to the two other 4GB sticks in the same two slots where the passed sticks were placed.  And so on.

 

Memtest86+ Guide-How To's - Overclockers Forums - http://www.overclockers.com/forums/showthread.php?t=409152

 

Tedious but that's the manner in which I would proceed.

 

Louis



#10 SleepyDude

SleepyDude

  • Malware Response Team
  • 2,995 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Portugal
  • Local time:01:59 AM

Posted 16 February 2017 - 10:32 AM

SIDE QUESTION: How many passes is 'normal' for one to assume RAM is OK? I'm fairly sure the RAM is OK as MemTest86+ made about 8 passes without a single error, and the uncorrectable memory alert is referencing different DIMMS, and it's not very liekly that there is something wrong with ALL of them AND MemTest86+ didn't find a problem with ANY of them.

 
Let the memory test run until you see "Pass complete..." similar to the following image:
4p1CobF4RFce7vjEZSd1GGaqg.png

Edited by SleepyDude, 16 February 2017 - 01:35 PM.

• Please do not PM me asking for support. Post on the forums instead it will increases the chances of getting help for your problem by one of us.
• Posts in the Malware section that are not replied to within 4 days will be closed. PM me or a moderator to reactivate.
• Please post your final results, good or bad. We like to know! Thank you!

 
Proud graduate of GeekU and member of UNITE
___
Rui

 
 


#11 rotoq

rotoq
  • Topic Starter

  • Members
  • 7 posts
  • OFFLINE
  •  

Posted 16 February 2017 - 10:34 AM

I'm half way through the second pass. I've to head out for a bit so no point in stopping now. I'll leave it on until I'm back.

Thanks for the advice. This isn't a criticism but a genuine question:

I thought memtest86 swept all the ram and performed tests on every 'block'. Assuming it doesn't find any errors, what advantage would there be in testing in pairs of dimms?

#12 hamluis

hamluis

    Moderator


  • Moderator
  • 55,550 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Killeen, TX
  • Local time:07:59 PM

Posted 16 February 2017 - 11:10 AM

I would do so...because my premise (and the recommendation by most) would be to test each module separately.  By testing more than 1 at a time, a flaw which could create an error...may be hidden by the fact that there are multiple sets of data being interpreted during testing.  There is the same possibility with the motherboard RAM slots.

 

I've only had one bad stick of RAM in my computer life.  When I tested it with the other module, no errors were evident.  When I moved to testing each module separately...it became clear that newer module which I had purchased...was defective.  I called Crucial, they gave me an RMA for the module...all was good with the new module.

 

Note that I did not say that's what you should do...I said I would pursue that procedure.  You do whatever you want...but you seem to have expended a lot of effort thus far with no results and a little more effort just might resolve your situation.  Then again, it may not.

 

Louis



#13 rotoq

rotoq
  • Topic Starter

  • Members
  • 7 posts
  • OFFLINE
  •  

Posted 16 February 2017 - 11:35 AM

Thanks for the reply, again I wasn't trying to be critical, merely curious ☺️

I'll add it my list of tests.

If I get this sorted and all my plex, samba, VMs etc back up I'll be elated. I didn't realise how much I relied on my home server until I replaced it with this one and it didn't work.

Lesson learned, don't get rid of the old one until it's replacement had proved its worth 😂😂

#14 hamluis

hamluis

    Moderator


  • Moderator
  • 55,550 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Killeen, TX
  • Local time:07:59 PM

Posted 16 February 2017 - 11:41 AM

Criticism...is one of the prime learning forces in the world today :), I don't take offense at curiosity or criticism :).

 

As long as I get a chance to respond :).

 

Keep us posted on your findings.

 

Louis


Edited by hamluis, 16 February 2017 - 11:42 AM.


#15 paul_s

paul_s

  • Members
  • 1 posts
  • OFFLINE
  •  
  • Local time:02:59 AM

Posted 17 August 2017 - 08:40 AM

Hi

 

Were you able to resolve this problem?  I'm having what appears to be exactly the same issue with a T5500.

 

Thanks






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users