WU failing / access violation errors in Firefox
2 Jul 2016 7:13:21 UTC
Topic 198672
(moderation:
Not able to complete WU in nearly two months now. All Gamma-ray pulsar binary search #1 v1.00 windows_intelx86 are failing.
Moreover, I'm getting access violation errors in Firefox that are invariably linked to Einstein WU's, i.e., either WU begin, resume from memory, have been crunching - for hours - or transitioned to another project leaving WU in memory where Einstein was crunching within the last 1/2 hour or so.
No other BOINC projects are experiencing prollems. No other Windows apps experience prollems. Browser does NOT experience any prollems despite having 300 tabs open except per aforementioned association with current, previous or initiation of Einstein WU. Rosetta may crash on rare occassions however, but EVERY time Einstein is associated per aforementioned description.
Windows does NOT manifest access violation bugchecks. I ran Prime95 for an hour with 0 warning / 0 errors.
WU failing / access violation errors in Firefox
)
Is this the same problem you were having here and here ?
You may have identified a bug in Windows Server 2003. Sadly Microsoft support ended on July 14, 2015.
RE: You may have identified
)
The repeated successful return of GRP work from these Server 2003 hosts would suggest otherwise:
Example 1
Example 2
Example 3
Example 4
The Professor's heritage machine (and my using that word is not a criticism) is most likely starved of memory...
300 tabs open in Firefox? Surely a typo but if not and assuming it is possible to open that many, trying to service the OS, run work from several projects and keep alive hundreds of web pages is asking an awful lot from 1.5GB of system memory and a PIII CPU. My limited understanding of access violation error is that an application attempts to write to memory address space that is already occupied by another app. (I do stand to be corrected on this).
Close those tabs in Firefox, free some memory and perhaps suspend your other projects and test the Einstein app again.
RE: RE: You may have
)
These are running on 64 bit processors i think, and only two appear to be running 32bit 2003 R2.
The error reported in two failed tasks are 0xc00002b5 (floating point exceptions) see https://msdn.microsoft.com/en-gb/library/cc704588.aspx - I expect FGRPB1 uses SSE.
Firefox just crashed. Guess
)
Firefox just crashed. Guess what? Einstein had begun running again - after five hours of previous crunching - where Firefox died 1 hour later.
Uptime = 13 Days 23 hrs
This is the FIRST occurrence of ANY prollem since I complained.
BTW: I'm waiting on Einstein WU crunch credits for the last WU I successfully uploaded after sompletion.
BTW BTW: this is the fist time SINCE I complained and SINCE I completed and uploaded my last WU that Firefox crashed. What's interesting is that Einstein keeps crunching; it never skipped a beat.
RE: RE: RE: You may
)
Is there any way to stop SSE+ WU from being downloaded?
I've suspended crunching of existing WU and no additional WU will be downloaded until I get credit for WU pending validation.
Firefox just crashed again.
)
Firefox just crashed again. Einstein WU restarted - previously left in memory - at 1811 EDT. Windows uptime: 16 days 18.3 hrs.
I've no clue how this can even be possibly related; the processes should be segregated from each other, no? And yet, Firefox NEVER crashes unless Einstein is implicated (per my previous description).
And just to be clear: since I started this thread, I've crunched all kinds of WU for various projects, and that they've restarted image memory resident - numerous times - is beyond reproach. I've received credit for nearly all - with the sole exception of my last Einstein WU pending validation - of them since I begat this thread.
So I DON'T KNOW; maybe I've got a signle bit that's bad in my 3x 512MB PC133 SDRAM. That'd be a shocker, eh?
AFAIK, Firefox crashes ALL THE TIME. I like it cause its SO MUCH BETTER TJAM IE8 which was what I used until web-sites began en masse non-support. Maybe I'll try out the Maxthon browser for a while.
Maybe my system is running out of stack or heap? Maybe whatever it IS using is leaves too fragmented free-mem for Firefox? If that was true, wouldn't I be failing validation of WU right & left?
Its ALWAYS Einstein that crashes Firefox v4x.
After reading
)
After reading this:
http://www.securityweek.com/maxthon-browser-sends-sensitive-data-china
I'm confuused if its worsr for the Chinise to look up my poop-chute, or Hillary's America?
AwlseIKnow is that Einstein causes Firefox to crash on my PIII 3x512MB PC133 SDRAM platformed Win2003 R2 system.
Maybe I'll use Chrome. Chrome should be good; NSA recommends Chrome.
I'm done. I had 26 days
)
I'm done.
I had 26 days uptime until I re-enabled Einstein.
There's something wrong with this app. No OTHER project gives me the grief that this one does.
I'm done.