These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Issues, Workarounds & Localization

 
  • Topic is locked indefinitely.
 

BSOD when running multiple accounts

Author
Toshiro Hasegawa
Blackwater USA Inc.
Pandemic Horde
#1 - 2014-09-20 01:24:34 UTC
I am not sure the best way to look into this issue.

It is new .. i have played for years without issue with 1-2-3-4 accounts at the same time.

In the last few months i have had several instances of PC lockup (video goes crazy, audio locks on one series of sounds) then BSOD.

Usually when one ship just gets a command.

I would like to look into it more .. i am an IT person so willing to get into the weeds on this one.

Its intermittent so dont know if i can reproduce.

Dont know if its driver or upset HD or whatnot .. but instead of beating down every path . thought id ask here how best to proceed to find and remedy the culprit.

What i find confusing is none of my hardware has changed .. no new software as of late either.

anyhow thanks in advance for any advice

History is the study of change.

Toshiro Hasegawa
Blackwater USA Inc.
Pandemic Horde
#2 - 2014-09-20 01:56:16 UTC
and just had third in a row .. was docked in both ships and just using the market ...

---
killed the audio and dropped graphix to min on 2nd account and its working now .. make me think it

History is the study of change.

GreenSeed
#3 - 2014-09-20 04:28:25 UTC
the software is not causing them, otherwise people would be going ape **** over it. also, hardware does degrade over time and specially when its used under heavy load all the time.

in order of likeliness of malfunction:

-RAM: have you ever had any BSOD with other RAM demanding applications? one eve is not demanding, but the more clients you have the more it will stress it.

it could be your RAM Dying, Windows has a memory check utility that's actually really good. ( Control Panel\System and Security\Administrative Tools -> Windows Memory Diagnostic on W7)

if the ram check returns with errors, then replace the sticks. if the PC is old, then finding replacements could be tricky, but not expensive.


-VGA: how old is your VGA? have you had any hangs on other GPU intensive applications? check the temperatures on it, if they are too high, doing a careful clean up of the Heat sink will lower the temp a few C. Dust, Dust everywhere. and if you have a cat or other house pet, after years of use, that probably looks like an H.R. Giger artwork.

if you notice a reduction on hangs after a clean up, then next step would be to buy some Thermal Conductivity Grease and CAREFULLY remove the heat sink from the vga, apply the lotion on its skin, and put it back together. the old grease will be tuned into a dry "clay like substance" by now so cleaning the surface of the GPU may be necessary, make sure you are grounded or you could damage it.

that should give more stability, but be warned that it has already sustained damage, and it will give up soon.

And lastly:

Check temperatures on your motherboard, specially on the NorthBridge. BUT BE CAREFUL NOT TO USE YOUR BARE HAND. it will burn you in a very nasty way if that's the problem. you can also check the back of the motherboard around the North Bridge, look for heat degradation/marks on the board itself.

if you don't know how to identify the North Bridge, it will always be between all the PCI slots and the CPU, its a big chip. almost 100% of them have at least some form of passive cooling. they also get some ventilation from the CPU cooler, so check if you don't have a family of cockroaches living there, blocking the cooling.

a basic diagram to know where to look would be:

PCI slots <----> North Bridge <----> CPU Slot
.......................................I
.......................................I
..South Bridge..........RAM Slots

(the south bridge is a smaller chip, will also have a heatsink, but will run much cooler)

if you run more than two clients at the same time, the load on the North Bridge will be more substantial that running one instance of an extremely demanding application. (Pc Architecture woes)

so chances are it will be extremely hot already. its not that much of a problem since they are built to run under lots of stress, but adding active cooling on it if you don't have any will help with stability.

just a small cooler will make a ton of difference. but the problem is, if the instability is caused due to north bridge heat, it will most likely die soon anyway. (the damage has been done.)


and finally, thee most likely scenario is that one of those parts, or all at the same time, have already given you years of eve and probably need to retire and become a media server of some kind.
Toshiro Hasegawa
Blackwater USA Inc.
Pandemic Horde
#4 - 2014-09-20 12:32:28 UTC
wow wasnt expecting someone to have to write so much ,, thanks for all that.

PC is a mix of new and old.

6950 is 3 years old
i7 4770 is one year old
Z89 board is one year old
3 x 4gig ddr 3 kingston is 3 years old
1 x 4gig ddr 3 kingston is 1 year old
HD is 3 years old

I will run cpu-z next time im playing with multiple accounts. see if i monitor any of the hardware...

no other BSOD .. only eve with multiple accounts running

i played for 1-2 hours last night with alt account grphx turned down with it spinning in station while main was in 1hr of a 50 ship brawl.

I need to look at the dump file .. but recall the last time i did that it took ages and wasnt very useful.

I will pop the hood and look at the GPU. might also start doing research on Toms Hardware for a replacement r9 280x maybe..

should probably also check the system logs. anyhow wife agro. . :P

History is the study of change.