0

Today my Dell T7400 suddenly started to beep at startup. I figured out is likely to be a memory issue.

I have used all eight (RAM ) slots (8x2GB). By removing / adding bar by bar I was able to restart the server by leaving the two last slots (at the bottom) empty. Then I have altered the RAM bars, but regardless of which one I use the issue (no start/beeping) appears as soon as I use the last slot (at the very bottom).

From that I can only conclude that it is not the RAM bar which is broken, but some problem with the the main board's last slot (is that correct?).

Q1: What would be the best way to find out the real root cause?

Q2: If I want to have 16GB again, can I use a configuration of 4x2GB and 2x4GB in the 6 operating slots.

Horst Walter
  • 159
  • 1
  • 1
  • 9
  • 1
    Try checking for grime/debris on the DIMM slot contacts. Also, did you determine what the POST code was based on the beeps? – user5071535 Aug 21 '15 at 19:30
  • 1
    Have you ruled out oxidization and contacts shorting? I would clean the problematic slot before trying anything else. – Alexander Obersht Aug 21 '15 at 19:31
  • Will check the slot again. but would a slot oxidize with the RAM bar in? It was OK yesterday, but failed today.Also I have altered and added / removed the bars several times today, which IMHO should also clean up the slot somehow. However, I will check it and thanks for the hint. – Horst Walter Aug 21 '15 at 19:47

2 Answers2

2

Check the DIMMs (ram bars) to see if they need to be paired and placed in a certain order. This is an older model (2007ish?) and those had finicky DDR2. Plus, ECC/registered memory itself is super picky. There literally may be a certain order and pairing the memory needs. I ran into this all the time with SC1425's. Something to do with the chipset as well.

Another note, Special notes: FB-DIMM Memory Risers are installed in the first four standard memory sockets only. Additional standard memory sockets then become unusable.

There's a pic in this forum link

http://en.community.dell.com/support-forums/desktop/f/3514/t/19392530

Here's the manual https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0CCYQFjABahUKEwi54NPogrvHAhVJGj4KHRJnAVs&url=http%3A%2F%2Fpdfstream.manualsonline.com%2F4%2F401888cd-1eec-44b5-8a52-a79cd245dedb.pdf&ei=BovXVbnFAsm0-AGSzoXYBQ&usg=AFQjCNEZtph5jrToXJnblj8fP1Z0XZBcIw&sig2=SfIFseVVcSeQVHJxHPsCCg

Fully Buffered DIMM (FBD) Memory Overview For optimal performance, fully buffered DIMMs (FBDs) should be installed in matched sets of four. This enables quad-channel operation and provides the highest memory bandwidth. The sets of four must be matched in size, rank, and organization. This information can normally be found on the FBD label. For example, the label of each FBD in a set of four might read: 1G 2R x8, in which 1G represents the size of the memory module, 2R represents the number of ranks, and x8 represents the organization. NOTICE: Full-length heat spreaders (FLHS) are required for all memory. Memory Installation Fully buffered DDR2 memory modules can also be installed in pairs of two; in this case, the computer will continue to operate, but with a slight reduction in performance. The pairs must be matched in size, rank, and organization. Install memory modules in order of their labels on the system board; matched sets of four should be installed first into DIMM slots 1-4 and then into DIMM slots 5-8 on the system board. Pairs of memory should be first installed into slots DIMM_1, and DIMM_2, then DIMM_3 and DIMM_4, and so on.

Mike Kauspedas
  • 56
  • 3
  • All very useful information, thank you: However, the server was ordered and configured from DELL and running fine for > 5 years (ordered 2009). I have never had any problems since today. For me that rules out a RAM misconfiguration. From what I can tell so far, that it seems to run fine with 6 bars, 7 bars gives an BIOS "mismatched RAM" error but still runs, 8 bars fail straight after switch on with beeps. No sign of dust or oxidation. So for me it looks like I simply have to accept the fact the boards is somehow broken. Thanks again for all the details above. – Horst Walter Aug 21 '15 at 21:29
  • 1
    Have you applied the latest BIOS and firmware updates from Dell? Had a similar issue in the past that turned out to be a BIOS update. Dell is actually pretty good at helping even after the warranty has expired, never hurts to open a chat to see how far you get. It's just odd that the slot would go bad and not the DIMM. – Mike Kauspedas Aug 22 '15 at 14:20
1

If the slot seems to be the problem you can visual inspect it. Check if it is not oxidized. Usually remove/insert of a memory module will clean the oxides.

Besides that the slots being broken, I would test for:

  • power supply - check the server's hardware log.
  • memory controller - which nowadays is part of CPU - swap CPUs to isolate the problem.
  • electromagnetic interference - check grounding.

Dell has a good server support, check if your server is still in warranty and open a ticket with Dell Support. You could also ask a question on linux-poweredge list, but check before the hardware log using DRAC or OMSA.

Mircea Vutcovici
  • 17,619
  • 4
  • 56
  • 83