I Broke it.

one of my rigs just ate 2 qmds. thats 900pts! pfft....

not sure what happened. not really mad, just a lil dissapointed.
im trying to see how high my subteam takes me, then i am going to go back folding for my favorite subteam (oc-rev.com) thats more of a long term plan though...

can anyone shed some light on it?

1. ran services.msc. - stopped both instances.
2. a few minuets later i put the machine in "standby" because i had a headace, and took a nice long nap in a very quiet room.
3. came out of standby a few hours later, restarted folding after a few minuets.
4. a few hours after that, i fired up EM III, and had a look, because i should have had at least one instance about ready to dump points, and another about halfway. im getting pretty good at estimating brogress between peeking at EM III - but this is showing 5 and 6 frames.
5. i do a happy dance, because they both dumped early. i check the stats, and i see nothing. nada. not even the 19 point "error credit"...
6. check the logs.
the cores deleted themselves. that means no credit at all.
ive never seen "bad frame checksum" before, but i know nothing out of the ordinary happened. what gives?

Loggage: (about the same for both)

15:03:06] Completed 1260 out of 2064 steps (61)
[15:03:06] Writing local files
[15:34:16] Completed 1280 out of 2064 steps (62)
[15:34:16] Writing local files


--- Opening Log file [September 18 00:42:27]


# Windows Console Edition #####################################################
###############################################################################

Folding@Home Client Version 5.02

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Program Files\FAH\FAH1
Executable: C:\Program Files\FAH\FAH1\FAH502-Console.exe
Arguments: -local -service -advmethods -verbosity 9 -forceasm

Warning:
(standard warning, edited)

[00:42:27] - Ask before connecting: No
[00:42:27] - User name: orionlion82 (Team 32)
[00:42:27] - User ID: 61F014610DC0C127
[00:42:27] - Machine ID: 1
[00:42:27]
[00:42:27] Loaded queue successfully.
[00:42:27] + Benchmarking ...
[00:42:30] The benchmark result is 6636
[00:42:30]
[00:42:30] + Processing work unit
[00:42:30] - Autosending finished units...
[00:42:30] Core required: FahCore_96.exe
[00:42:30] Trying to send all finished work units
[00:42:30] + No unsent completed units remaining.
[00:42:30] - Autosend completed
[00:42:30] Core found.
[00:42:30] Working on Unit 09 [September 18 00:42:30]
[00:42:30] + Working ...
[00:42:30] - Calling 'FahCore_96.exe -dir work/ -suffix 09 -checkpoint 30 -service -forceasm -verbose -lifeline 3896 -version 502'

[00:42:30]
[00:42:30] *------------------------------*
[00:42:30] Folding@Home QMD Core
[00:42:30] Version 1.04 (Apr 7, 2005)
[00:42:30]
[00:42:30] Preparing to commence simulation
[00:42:30] - Assembly optimizations manually forced on.
[00:42:30] - Not checking prior termination.
[00:42:31] - Expanded 260973 -> 826210 (decompressed 316.5 percent)
[00:42:31]
[00:42:31] Project: 1912 (Run 64, Clone 6, Gen 46)
[00:42:31]
[00:42:31] Writing local files
[00:42:31] Extra SSE2 boost OK.
[00:42:31] Entering QMD...
[00:43:24] System: p1912_ALA-dipeptide_umb2
[00:43:24]
[00:43:24] Verifying checksum
[00:43:24]
[00:43:24] Folding@home Core Shutdown: BAD_FRAME_CHECKSUM
[00:43:26] CoreStatus = 70 (112)
[00:43:26] + The core could not validate the current work unit for processing.
[00:43:26] Deleting current work unit & continuing...
[00:43:30] Trying to send all finished work units
[00:43:30] + No unsent completed units remaining.
[00:43:30] - Preparing to get new work unit...
[00:43:30] + Attempting to get work packet
[00:43:30] - Will indicate memory of 2046 MB.
[00:43:30] - Connecting to assignment server
[00:43:30] Connecting to http://assign.stanford.edu:8080/
[00:43:31] Posted data.
[00:43:31] Initial: 40AB; - Successful: assigned to (171.64.122.113).
[00:43:31] + News From Folding@Home: Welcome to Folding@Home
[00:43:31] Loaded queue successfully.
[00:43:31] Connecting to http://171.64.122.113:8080/
[00:43:32] Posted data.
[00:43:32] Initial: 0000; - Receiving payload (expected size: 261485)
[00:43:38] - Downloaded at ~42 kB/s
[00:43:38] - Averaged speed for that direction ~37 kB/s
[00:43:38] + Received work.
[00:43:38] + Closed connections
[00:43:43]
[00:43:43] + Processing work unit
[00:43:43] Core required: FahCore_96.exe
[00:43:43] Core found.
[00:43:43] Working on Unit 00 [September 18 00:43:43]
[00:43:43] + Working ...
[00:43:43] - Calling 'FahCore_96.exe -dir work/ -suffix 00 -checkpoint 30 -service -forceasm -verbose -lifeline 3896 -version 502'

[00:43:43]
[00:43:43] *------------------------------*
[00:43:43] Folding@Home QMD Core
[00:43:43] Version 1.04 (Apr 7, 2005)
[00:43:43]
[00:43:43] Preparing to commence simulation
[00:43:43] - Assembly optimizations manually forced on.
[00:43:43] - Not checking prior termination.
[00:43:44] - Expanded 260973 -> 826210 (decompressed 316.5 percent)
[00:43:44]
[00:43:44] Project: 1912 (Run 64, Clone 6, Gen 46)
[00:43:44]
[00:43:44] Writing local files
[00:43:44] Extra SSE2 boost OK.
[00:43:44] Entering QMD...
[00:44:37] System: p1912_ALA-dipeptide_umb2
[00:44:37]
[00:44:37] Performing initial WF calculations
[00:44:37] - Number of total steps will change until convergence
[00:46:27] Completed 0 out of 2000 steps (0)
[01:12:25] Completed 21 out of 2021 steps (1)
[01:12:25] Writing local files
[01:37:30] Completed 41 out of 2041 steps (2)
[01:37:30] Writing local files
[02:04:17] Completed 62 out of 2062 steps (3)
[02:04:17] Writing local files
[02:07:16] Timered checkpoint triggered.
[02:07:22] WF converged, jumping to MD

i have no idea what could have caused this. its a bummer.

Standbye usually messes up my wu's perhaps it did the same to you.

Steveo989 said:

Standbye usually messes up my wu's perhaps it did the same to you.

Click to expand...

thats news to me. the "all while you use your computer normally" part... hmmph!

i mean i can see how it might have that potential,
shouldnt do that IMHO,
i even had the dumb luck to turn off the service before, and restart it after. if anything that should have prevented suspend from being a part of the equation at all...
anyone else have bad experiences with power management?

not sure why you even worried about puting it on standby. there is realy no point, just kill the monitor, and hit the sack, or whatever

yeah, thats what i usually do, but i had me a proper headache this time around, and thought the quiet factor would be nice (it was) not that my rig is all that loud, but im sure a few folders use standby every now and then.

i dunno, i guess this might be the issue, and its doubtful that stanford knows, or cares.
might as well just shut up and live with it i suppose. blah.

orion, i think i may know how to help you.

right click on your desktop, poperties, screensaver tab, power button inside of screensaver tab, change powerscheme to laptop.

next, go to your client.cfg for the F@H, (which should look similar to this:

[settings]
username=F(OC)LD
team=32
asknet=no
machineid=1
local=438
bigpackets=yes

[http]
active=no
host=localhost
port=8080
usereg=no

[clienttype]
type=1

[core]
priority=0
cpuusage=100
disableassembly=no
checkpoint=30
ignoredeadlines=no

[power]
battery=no

Click to expand...

in the last section, change the battery=no to battery equals yes.

now, whenever yer system hibernates (this doesn't always work, but i've had about 50/50 success with this method), it thinks it is a laptop, and will pause the F@H process, and should allow it to continue upon 'reanimation'.

i dont' know if this will fix your problem, but its a possible solution.

THF!
youre an effing genious man! i absolutely cannot believe it!
thats the slickest trick ive ever seen!
WOW!
oh, man, so giddy....
(catches breath)
youre my hero for the day!
gonna get some chow, and give it a shot!

Standby? What is this standby you speak of?

Standby = the most useless use for a PC.

you are right shard, but once upon a time, standby was usefull

awww, bummer!

i just gave it a shot just for fun i hit the suspend button on the keyboard, had a sandwich -

another QMD gone, this time it E_U_E'd...

the other one i had running was fine though.
as promised, a 50% death rate.

now 1350 behind! ughhh!
and i was hoping to spank alabama cajun this week, once and for all...

Not the most brilliant move bud. So from now on, how about just NOT hitting that magical button, ok?

lol... so a 50% rate for you as well? at least i know it can work if you need it to... if you get lucky

orionlion82 said:

and i was hoping to spank alabama cajun this week, once and for all...

Click to expand...

Sorry to hear about your problems.

BTW, check your 'Threats' a page back........