UNSOLVED
R730 with 2x E5-2699v + 512GB - Multi-bit Memory Error in differentiating slots
We have an R730 that we're putting 2x E5-2699v4 processor in (SR2JS), along with 512GB RAM (16x32GB DDR4 PC4-2400T) (2x 1100W PSU's)
We have updated the BIOS to the latest (2.3.4).
_____________________________________________________________________
When using 1 CPU in the server, and 8 DIMMs, the server boots up fine.
Swapping the CPU1 for CPU2, with the same configurations (1 CPU + 8x 32GB DMIMs), boots fine as well, so it somewhat tells me its not the CPU's and the memory we're using that are the problem.
______________________________________________________________________
If i then take the second CPU, and put it in socket 2 (so 2 CPU's are in use), and use 4 of the 8 DMMs from the first CPU testing setup (so I know they work), into the second CPU DIMM slots..
so: (CPU1 + 4 DIMMs then CPU2 + 4 DIMMs)
The server then boots up with a multi-bit memory error in DIMM_B2. If we boot again, the memory error goes to B1 and then alternates sporadically.
________________________________________________________________________
Sometimes the R730 will boot all the way to be able to get to OpenManage with this configuration (Eventually after the multi-bit error occurs on the LCD screen, the server tries to continually configure itself and finds a configuration good enough to boot), but in OpenManage, it shows B1 & B2 DIMM Slots as having no memory there, which I suspect it disables to help it to still boot?
Mind you, these errors (B1 and B2) are coming from DIMM slots that have memory in them that was just tested 'Good' while in the server when it had the single CPU setup + 8 DIMMs (so I don't believe it to be the memory DIMMs that are bad since they work fine with 1 CPU.)
___________________________________________________________________________
I also tested some DDR4-PC4-2133p RAM, and it works good until we start to populate multiple 4 or more slots, it throws and error.
___________________________________________________________________________
All firmware is updated and BIOS updated.
It has an H730 RAID Controller in it.
___________________________________________________________________________
What am I missing? On Dell's website configurator, the option to choose 2x E5-2699v4 processors, plus 16 x 32GB 2400T DIMMs is an option there, so this set up should work. Plus I saw a review on a storage website, where they used this same set up with no issues (aside from having to update the BIOS)
___________________________________________________________________________
I originally thought it was a motherboard issue, but we put 2x E5-2620v4 processors in the machine, with all of the same 512GB 2400T of RAM (16 x 32GB), and it operated just fine. So it took the lower end processors with no issues.
____________________________________________________________________________
If we put 2x E5-2699v4 processors in there with all 512GB, it throws a Mutli-Bit error, and then tries to continually configure the memory, but eventually just shuts down.
This system isn't covered under Dell Warranty unfortunately as its from Canada. I try to search the Service Tag on their site and nothing comes up, so Calling Dell isn't really an option for me.
Any help you can provide or insight, I would greatly appreciated
They are supported DIMMs.
32GB 2Rx4 DDR4 2400T
I believe that falls in line with what they have in their Owners Manual and Tech Guide. (One of the is outdated)
http://topics-cdn.dell.com/pdf/poweredge-r730_Owner's%20Manual_en-us.pdf
Page 94 on this PDF shows it to be supported as well. We are populating the slots properly as it shows there as well.
I'm wondering maybe we have to populate only a couple chips at a time, then power down, put in more, and boot again. Repeat until all is installed? Just a guess as I'm not sure if that's even something, but I"ll give it a shot on Monday.
I also saw in this Owners Manual PDF on page 88 it says something about different Memory Modes, (Performance Optimized, Custom, or Dense Configuration Optimized)
Does anyone know if that's something that needs to be done for larger memory builds like this? I didn't see any more information on it in the PDF.theflash1932
11 Legend
•
16255 Posts
2215
0
Posted January 7th, 2017 18:00
The one at a time thing? No, I've never seen that, on any machine. You should be using Optimized Mode (for now).
What is the make and model of the memory modules you are using?
I thought I'd let people know what caused this issue.
Even though each CPU would work in a Single CPU configuration, it was a bad CPU that cause the errors when using both in a Dual CPU configuration.
Ordered a replacement CPU and put it in without any issue.
The fact it gave off a Multi-Bit Error in certain DIMM slots, but the errors didn't follow with the CPU when swapped, was what really threw me off.
ComputerWerks Inc
1 Message
2215
0
Posted November 17th, 2017 13:00
How did you determine which CPU it was before ordering it or did you just order one to troubleshoot?
I don't fully recall as we build so many systems, however I believe we did just bring in a third CPU, and test it after that and it worked.
Since these CPU's have Memory Controllers in them, it can give different errors that don't always point to the CPU (DIMM slot errors) so I believe we brought in a new one and tested it from there.
P
ECSunix
PortalNET2
1 Rookie
•
72 Posts
0
0
Posted September 17th, 2024 20:25
Hiya..
thanks i am having the same issue with 2 CPUs E5-2699v4 SR2JS on a Dell R630.. i even tought was Memory slot in hardware motherboard damaged and bought another r630 chassi.. and it throws out same error.. funny enought with older E5-2640v3 cpu it booted fine, will now test single core cpu on cpu1 to test if it boots.. and will test second cpu on second R630 to see if it boots also .. even if they do boot ok as single cpu on both machines.. why will they throw error when combined together on same server? i tested both cpus on both servers and had the exact same error on Memory DIMM Slot A1. allways throws error on first slot memory.. if i press F1 it boots normally but i loose 32GB ram one 16gb on A1 slot and also 16GB on B1 slot from cpu2 ..which is annoying
DELL-Joey C
Moderator
•
4170 Posts
•
20965 Points
0
0
Posted September 18th, 2024 03:51
Hi,
I've responded to your new post on the issue here: https://dell.to/4e6yZVp. I'll leave this post for the public to response to you if they have any other feedback than me.
DELL-Joey C
Social Media and Communities Professional
Dell Technologies | Enterprise Support Services
#IWork4Dell
Did I answer your query? Please click on ‘Mark as Accepted Answer’ if I did.
theflash1932
11 Legend
•
16255 Posts
2215
0
Posted January 7th, 2017 14:00
Sure it should work, but are you using memory with validated specs? CL, clock speed, layout, rank, etc.? ALL could be important and could cause some but not all memory to initialize. System requirements for RAM will depend on the Intel specs for the controller AND the Dell specs for the specific memory configuration. Just because online shows 32GB DIMM's will work does not mean that ALL 32GB DIMM's will work and/or in all configurations. Check the compatibility.