I hate 1808!

VeNT

VeNT

VeNT

Soldato
Joined
9 Jan 2003
Posts
21,018
Location
Cornwall
ok, I've got ANOTHER sodding 1808 , but this REALY takes the biscuit.
FAHMon tells me that I downloaded this WU at 17:48 on the 8th of aug, that I've only done 10% of the WU and that its going to take another 35 days 21 hours 40 mins (and rising) at this rate.
now I've just checked that rig (3200 XP, 1gig ram) FahCore_82.exe IS using 99% of the CPU, BUT System Idle Process seems to have 10-20x the CPU time :confused:
I just don't get it! if the core is getting 99% of the CPU time, how can SIP have MORE hours logged (and its way more, the core is showing 40 mins where as SIP has hours logged)
I've rebooted the client to see if that helps, but I dunno, I may have to re-install FAH on here :mad:
 

MGP

MGP

MGP

Soldato
Joined
24 Oct 2004
Posts
2,584
Location
Surrey
Yep something weird on my latest 1808 too. Currently shows as complete in FAHMon, yet I only got it, I think, within the last hour, and my CPU is flat out too.

*Perhaps I should go back to Boinc :eek:

edit: Just rechecked. FAHMon is in fact reporting the status of the last WU, which was a 1808 as complete, and has ignored the fact that that WU seems to have uploaded (my credit went up) and is paying no attention to whatever the PC has been up to for the last hour. Rebooting the PC made no difference.
 
Last edited:

rich99million

rich99million

rich99million

Soldato
Joined
26 Dec 2002
Posts
9,348
Location
Derbyshire
The ETA of 10 days now will be based on the rubbish previous performance so is obviously going to be way out. Sounds like it's running fine now so unlikely that the core/client or WU is corrupt in any way.

MGP - What is the FAHlog showing? It's not unknown for the client to stop reporting progress properly whilst it's still actually working and making progress. However a restart should have sorted that - is it possible you have a new core or WU that FahMon doesn't recognise (or even an older version of FahMon?)
 

MGP

MGP

MGP

Soldato
Joined
24 Oct 2004
Posts
2,584
Location
Surrey
Looks like it's reporting correctly now, must have just taken a while for the hamsters to turn the wheel, devoting a bit of energy to FAHMon rather than WU crunching.

Running FAHMon 2.1.1 which is I presume the latest one since I only defected to origami last week.
 

VeNT

VeNT

VeNT

Soldato
OP
Joined
9 Jan 2003
Posts
21,018
Location
Cornwall
rich99million said:
The ETA of 10 days now will be based on the rubbish previous performance so is obviously going to be way out. Sounds like it's running fine now so unlikely that the core/client or WU is corrupt in any way.

MGP - What is the FAHlog showing? It's not unknown for the client to stop reporting progress properly whilst it's still actually working and making progress. However a restart should have sorted that - is it possible you have a new core or WU that FahMon doesn't recognise (or even an older version of FahMon?)
Click to expand...

there was about 4-5 pages (or more) of checkpoints (15 mins) between two steps when I was having that uber long ETA
its down to 2 days now with 1% an hour atm (strangly slow?)

Project : 1808
Core : Amber
Frames : 100
Credit : 153

-- Hal 9000 --

Min. Time / Frame : 13mn 58s - 157.75 ppd
Avg. Time / Frame : 52mn 00s - 42.37 ppd

-- Brenda --

Min. Time / Frame : 14mn 09s - 155.70 ppd
Avg. Time / Frame : 14mn 41s - 150.05 ppd


-- Bot 001 --

Min. Time / Frame : 14mn 25s - 152.82 ppd
Avg. Time / Frame : 14mn 33s - 151.42 ppd

Hal is the one with the issue as you can see, 3200 1gig
brenda is a 3500 Athlon 64 with 1gig
Bot 001 is a Athlon XP 2500 barton with 512.

just goes to show how little difference there is somtimes.
 
Last edited:

rich99million

rich99million

rich99million

Soldato
Joined
26 Dec 2002
Posts
9,348
Location
Derbyshire
I've had WUs stall like that before but only when another higher priority process has gone awry and nicked all the CPU time itself (my USB wireless adapter used to do it - turned out to be some form of virus) like you said they just kept checkpointing but the core wasn't getting any CPU time.

In case it's the core itself you could stop the client, delete the offending core, then restart the client and it'll download a fresh copy. The same goes for the client - you can just download a fesh copy and overwrite the current one. If it's neither of those and you're fed up of this WU then delete the Work folder and the queue.dat file and restart the client to make it download a new WU (which may or may not be the same as the one you just got)

However I fear this sounds like something that could be windows messing about - probably worth doing a thorough scan to make sure you've not got any nasties
 

VeNT

VeNT

VeNT

Soldato
OP
Joined
9 Jan 2003
Posts
21,018
Location
Cornwall
rich99million said:
I've had WUs stall like that before but only when another higher priority process has gone awry and nicked all the CPU time itself (my USB wireless adapter used to do it - turned out to be some form of virus) like you said they just kept checkpointing but the core wasn't getting any CPU time.

In case it's the core itself you could stop the client, delete the offending core, then restart the client and it'll download a fresh copy. The same goes for the client - you can just download a fesh copy and overwrite the current one. If it's neither of those and you're fed up of this WU then delete the Work folder and the queue.dat file and restart the client to make it download a new WU (which may or may not be the same as the one you just got)

However I fear this sounds like something that could be windows messing about - probably worth doing a thorough scan to make sure you've not got any nasties
Click to expand...
nod 32s not showing anything
I think it was just playing up as it seems ok now.
 

Steevo38

Steevo38

Steevo38

Associate
Joined
23 Oct 2002
Posts
708
Location
Colchester
I'm no fan of these wu's either.

These 1808 ambers are the only ones that give similar ppd on my opty as on my E6300.

1809's are about 15% faster on my opty. Very strange as all other wu's are at least 30% faster on the Intel???


1478's are my favourites. :D

Project : 1478
Core : Gromacs
Frames : 100
Credit : 364


-- core2duo1 --

Min. Time / Frame : 4mn 59s - 1051.83 ppd
Avg. Time / Frame : 4mn 59s - 1051.83 ppd


-- core2duo2 --

Min. Time / Frame : 5mn 06s - 1027.76 ppd
Avg. Time / Frame : 5mn 06s - 1027.76 ppd


Fold on team 10
 

shadowscotland

shadowscotland

shadowscotland

Soldato
Joined
31 May 2006
Posts
7,564
Location
West London
These are my best so far - running both at moment so almost 600 ppd

Project : 2414
Core : Gromacs
Frames : 100
Credit : 600


-- thread 2 --

Min. Time / Frame : 29mn 20s - 294.55 ppd
Avg. Time / Frame : 29mn 29s - 293.05 ppd

Project : 2409
Core : Gromacs
Frames : 100
Credit : 600


-- thread 5 --

Min. Time / Frame : 29mn 54s - 288.96 ppd
Avg. Time / Frame : 31mn 58s - 270.28 ppd
 
You must log in or register to reply here.
Back
Top Bottom