EEVblog > News/Suggestions/Help

Server down 13th August

<< < (2/2)

EEVblog:
It's failed AGAIN!  :palm:

EEVblog:
Reported error

rsjsouza:
These situations always remind me of this classic:

thm_w:

--- Quote from: Ian.M on August 14, 2021, 07:51:35 am ---WTF??? If the memory was compromised due to the April 4th fire and water damage event, the whole motherboard will also be compromised.  It smells like they cant diagnose the failure so are 'shotgunning' it with unnecessary parts replacement.  Be glad its their box you are renting!

OTOH, refitting heatsinks with fresh thermal paste is the proper thing to do if there is any possibility that a system has been handled roughly enough to disturb them.

--- End quote ---

Isn't the whole point of server gear you just swap out everything, unless you are 100% certain of the problem?
Then test it outside on your own time if you really want to reuse it. Possibly for more than a few hours, I've had memtest issues show up at the 3hr+ mark.

Ian.M:

--- Quote from: thm_w on August 16, 2021, 08:44:15 pm ---Isn't the whole point of server gear you just swap out everything, unless you are 100% certain of the problem?
Then test it outside on your own time if you really want to reuse it. Possibly for more than a few hours, I've had memtest issues show up at the 3hr+ mark.

--- End quote ---
As you point out, full replacement would have been the best option for everything, but we know GorillaServers are 'bargain-basement', so 'swap out everything' probably isn't in their 'game-plan' unless a large customer is holding their nuts to the fire!

My point was that any mishap that could damage previously functioning memory would be equally likely to damage the motherboard.  OTOH if they were total muppets and removed all the RAM to clean it in bulk without appropriate ESD precautions they *could* be seeing a higher than normal incidence of (self-inflicted) RAM failures.

Navigation

[0] Message Index

[*] Previous page