November 13, 20232 yr Hello, I been using unraid for about a year or two now, and after upgrading my ram to 64GB I have been having an error where my whole system would crash and then reboot immediately after about 20 hours or so of being on consistently. I also have the "Machine Check Events detected on your server" error on my system as well. Not sure how to fix this either. I have attached a copy of my diagnostics file. So I'm just not sure where to start to fix the problem outside of running the ram test. Which I ran for an hour and a half and had zero errors. Thank you tower-diagnostics-20231113-0459.zip Edited November 13, 20232 yr by Mrtj18
November 13, 20232 yr Community Expert 1 hour and a half for memtest is not much, but you can also try removing the extra RAM and se if the server returns to stable.
November 13, 20232 yr we talking commodity memory, or ECC? if the latter, run the following... grep "[0-9]" /sys/devices/system/edac/mc/mc*/csrow*/ch*_ce_count it'll show you if any chips have been having correctable faults. Edited November 13, 20232 yr by sota
November 13, 20232 yr Author 52 minutes ago, JorgeB said: 1 hour and a half for memtest is not much, but you can also try removing the extra RAM and se if the server returns to stable. I have the test running currently while I go off to work, so hopefully after 9 hours I see some different results Also I have 2 sticks of 32GB of ddr4 RAM g. Skill brand. Non eec. So you recommend I try one stick at a time to determine which one could be at fault?
November 13, 20232 yr Author 54 minutes ago, sota said: we talking commodity memory, or ECC? if the latter, run the following... grep "[0-9]" /sys/devices/system/edac/mc/mc*/csrow*/ch*_ce_count it'll show you if any chips have been having correctable faults. I have non-error correcting memory, can I run this command on my system still?
November 13, 20232 yr Community Expert 10 minutes ago, Mrtj18 said: So you recommend I try one stick at a time to determine which one could be at fault? Worth a try.
November 13, 20232 yr Author 5 minutes ago, JorgeB said: Worth a try. Okay cool, and how long would you recommend, I let each memory test run?
November 13, 20232 yr Community Expert Try running the server normally, with just one stick, if it crashes try the other one, that would basically rule out bad RAM.
November 13, 20232 yr Author 11 minutes ago, JorgeB said: Try running the server normally, with just one stick, if it crashes try the other one, that would basically rule out bad RAM. Will try this evening, thanks!
November 13, 20232 yr 3 hours ago, Mrtj18 said: I have non-error correcting memory, can I run this command on my system still? I have no idea, as the only machine I run unRAID on has ECC
November 14, 20232 yr Author Solution Update, so after running the ram test for 11 hours with no errors I decided to open up my case and double check if everything was plugged in correctly, double check my bios and even move the ram to the other 2 slots on my motherboard. Upon opening my case I noticed that I had one pcie cable split into two 8 pin connections plugged into my rtx 3080 gpu at once. Which I think caused the random reboots because the GPU was not getting enough power. Once I plugged the rtx 3080 into different 8 pin cables, I have been running my system for 11 hours no errors ( it was crashing once every 6-7 hours )with both ram sticks in. So i will mark this as solved and chalk it up to user setup error. due to GPU pins lol. Thanks for all your help!
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.