FAHClient V7.1.52 released (1st Public-Release)

[Miller855]

snip
with previous versions i could just use Command C & Command V to copy and paste my Passkey

As far as I know, Command-C and Command-V have never worked as copy and paste in FAHControl on OS X, and there is an open ticket on that. The Linux and Windows style Control-C, Control=V and Control-A key combos work as cut, paste and select all.
snip

until V 7, I used increase to monitor my folding, it always let me copy & paste.....
I learned something new today, thanks !

Miller

[Jesse_V]

Fun fact: the "rollout" of this client is a single day early for the anniversary of the release date of F@h on the PS3 in 2008 and I think it's also really close to the anniversary of when Google co-founder Sergey Brin put F@H on the Google Toolbar in 2002.

I like how F@h's been doing all this science, earning all these awards, and smashing all these records while the high-performance clients are just now starting to come out from beta! Not there yet, but just wait, it's going to be epic...

[Stonecold]

Hi,

I've tried running FAHViewer on Ubuntu 12.04b1 but keep getting this error:

Code: Select all

FAHViewer: error while loading shared libraries: libGLEW.so.1.7: cannot open shared object file: No such file or directory

There is no libglew1.7 on Ubuntu; the latest in the repositories is 1.6.

I'm having the exact same problem on Kubuntu 11.10 when I try running it. When I try starting FAHViewer from FAHControl, nothing happens. When I try to run it from Terminal, I get this:

Code: Select all

alex@ubuntu:~$ cd /usr/bin
alex@ubuntu:/usr/bin$ ./FAHViewer
./FAHViewer: error while loading shared libraries: libGLEW.so.1.7: cannot open shared object file: No such file or directory

So I think this is a problem with the Linux releases (or at least Ubuntu), because I can see more people than me are having this problem.

[vmzy]

Sometimes next-unit-percentage didnt work properly .

The first WU(Project: 7008 (Run 2, Clone 13, Gen 40)) is correct .

Code: Select all

03:48:50:WU00:FS00:0xa4:Completed 9900000 out of 10000000 steps  (99%)
03:48:51:WU01:FS00:Connecting to assign3.stanford.edu:8080

But the second one(Project: 8012 (Run 159, Clone 6, Gen 2)) was wrong .

Code: Select all

14:06:54:WU01:FS00:0xa4:Completed 247500 out of 250000 steps  (99%)
14:11:49:WU01:FS00:0xa4:Completed 250000 out of 250000 steps  (100%)
14:11:51:WU00:FS00:Connecting to assign3.stanford.edu:8080

Does this bug related to Project?

BTW:default next-unit-percentage is 99 or 98?In this page it says 98(https://fah-web.stanford.edu/projects/F ... ncesV6ToV7).

2012-4-20 update

Code: Select all

04:53:22:WU00:FS00:0xa4:Project: 8004 (Run 58, Clone 23, Gen 107)
09:04:13:WU00:FS00:0xa4:Completed 247500 out of 250000 steps  (99%)
09:06:52:WU00:FS00:0xa4:Completed 250000 out of 250000 steps  (100%)
09:06:52:WU00:FS00:0xa4:DynamicWrapper: Finished Work Unit: sleep=10000
09:06:53:WU01:FS00:Connecting to assign3.stanford.edu:8080
09:06:58:WU01:FS00:News: Welcome to Folding@Home

All 80xx project on my machine were missing next-unit-percentage .

[bruce]

It appears that development thinks this bug is fixed. See Ticket #842

AFAIK, p8012 writes a checkpoint every 15 minutes, so there may be no report between 14:06:54 and 14:11:49.

[vmzy]

checkpoint didnt work properly .
I have set checkpoint interval to 3 min.But it have no effect .I wiill paste restart log below, You can see the start step was wrong .

2012-04-04 shutdown log

Code: Select all

15:42:25:WU00:FS00:0xa4:Completed 400000 out of 10000000 steps  (4%)
15:50:32:Lost lifeline PID 5608, exiting
15:50:33:Server connection id=1 ended
15:50:33:FS00:Shutting core down
15:50:38:WU00:FS00:0xa4:Client no longer detected. Shutting down core
15:50:38:WU00:FS00:0xa4:
15:50:38:WU00:FS00:0xa4:Folding@home Core Shutdown: CLIENT_DIED
15:50:38:Clean exit

2012-04-05 startup log(there is no checkpoint between 15:42:25 and 15:50:32)

Code: Select all

01:11:15:WU00:FS00:0xa4:Project: 7008 (Run 2, Clone 13, Gen 40)
01:11:15:WU00:FS00:0xa4:
01:11:15:WU00:FS00:0xa4:Assembly optimizations on if available.
01:11:15:WU00:FS00:0xa4:Entering M.D.
01:11:21:WU00:FS00:0xa4:Using Gromacs checkpoints
01:11:21:WU00:FS00:0xa4:Mapping NT from 4 to 4
01:11:22:WU00:FS00:0xa4:Resuming from checkpoint
01:11:22:WU00:FS00:0xa4:Verified 00/wudata_01.log
01:11:22:WU00:FS00:0xa4:Verified 00/wudata_01.trr
01:11:22:WU00:FS00:0xa4:Verified 00/wudata_01.xtc
01:11:22:WU00:FS00:0xa4:Verified 00/wudata_01.edr
01:11:22:WU00:FS00:0xa4:Completed 400001 out of 10000000 steps  (4%)

2012-04-05 shutdown log

Code: Select all

15:43:53:WU00:FS00:0xa4:Completed 8300000 out of 10000000 steps  (83%)
15:51:52:Server connection id=1 ended
15:51:52:Lost lifeline PID 1152, exiting
15:51:53:FS00:Shutting core down
15:51:56:WU00:FS00:0xa4:Client no longer detected. Shutting down core
15:51:56:WU00:FS00:0xa4:
15:51:56:WU00:FS00:0xa4:Folding@home Core Shutdown: CLIENT_DIED
15:51:56:Clean exit

2012-04-06 startup log(there is no checkpoint between 15:43:53 and 15:51:52)

Code: Select all

01:09:46:WU00:FS00:0xa4:Project: 7008 (Run 2, Clone 13, Gen 40)
01:09:46:WU00:FS00:0xa4:
01:09:46:WU00:FS00:0xa4:Assembly optimizations on if available.
01:09:46:WU00:FS00:0xa4:Entering M.D.
01:09:49:Server connection id=1 on 0.0.0.0:36330 from 127.0.0.1
01:09:52:WU00:FS00:0xa4:Using Gromacs checkpoints
01:09:52:WU00:FS00:0xa4:Mapping NT from 4 to 4
01:09:52:WU00:FS00:0xa4:Resuming from checkpoint
01:09:52:WU00:FS00:0xa4:Verified 00/wudata_01.log
01:09:53:WU00:FS00:0xa4:Verified 00/wudata_01.trr
01:09:53:WU00:FS00:0xa4:Verified 00/wudata_01.xtc
01:09:53:WU00:FS00:0xa4:Verified 00/wudata_01.edr
01:09:53:WU00:FS00:0xa4:Completed 8300001 out of 10000000 steps  (83%)

2012-04-06 shutdown log

Code: Select all

16:20:10:WU00:FS00:0xa4:Completed 45000 out of 1500000 steps  (3%)
16:33:17:Lost lifeline PID 1692, exiting
16:33:18:FS00:Shutting core down
16:33:18:Server connection id=1 ended
16:33:22:Clean exit
16:33:22:WU00:FS00:0xa4:Client no longer detected. Shutting down core
16:33:22:WU00:FS00:0xa4:
16:33:22:WU00:FS00:0xa4:Folding@home Core Shutdown: CLIENT_DIED


2012-04-07 startup log(checkpoint seems correct )

Code: Select all

01:41:55:WU00:FS00:0xa4:Project: 7809 (Run 8, Clone 205, Gen 51)
01:41:55:WU00:FS00:0xa4:
01:41:55:WU00:FS00:0xa4:Assembly optimizations on if available.
01:41:55:WU00:FS00:0xa4:Entering M.D.
01:42:01:WU00:FS00:0xa4:Using Gromacs checkpoints
01:42:01:WU00:FS00:0xa4:Mapping NT from 4 to 4
01:42:03:WU00:FS00:0xa4:Resuming from checkpoint
01:42:03:WU00:FS00:0xa4:Verified 00/wudata_01.log
01:42:03:WU00:FS00:0xa4:Verified 00/wudata_01.trr
01:42:03:WU00:FS00:0xa4:Verified 00/wudata_01.xtc
01:42:03:WU00:FS00:0xa4:Verified 00/wudata_01.edr
01:42:04:WU00:FS00:0xa4:Completed 49890 out of 1500000 steps  (3%)

you can see .
Project: 7008 (Run 2, Clone 13, Gen 40) will Shutting down core first ,then Clean exit, and it checkpoint was wrong .It only checkpoint at percent finished .
Project: 7809 (Run 8, Clone 205, Gen 51)will Clean exit first ,then Shutting down core , and it checkpoint was correct.

BTW:I add -forceasm parameter on my V7 ,Does it cause this bug .

[bruce]

@ vmzy

The parameter -forceasm is ignored by FahCore_a4.

I have confirmed that setting checkpoint interval does work with FahCore_a4. After you change that setting, it takes effect when the NEXT Wu starts (though you can pause and restart the current WU if you need it to happen quickly.

We don't recommend changing that setting. If it's too long, you have to repeat more work when work on a WU is resumed after some kind of pause. If it's too short, you increase the chances of a corrupt WU due to interrupting progress while a new checkpoint is still being written to disk.

[vmzy]

But I had set 3min checkpoint 5 months ago.And had restart my fah 100+ times.Why it took no effect ?
I check history log(5 months) just now .I noticed all 70xx project ignore checkpoint setting(only have percent checkpoint) .Maybe this is not bug of client or core .Just a bug of 70xx project.

[filu]

I think that the client would do well to implement the notification of updates. Users as quickly and efficiently they could pass on the newer versions of the client.

[7im]

Update notifications is already a documented feature request.

[robert2223]

OK, downloaded 7.1.52. Then found the latest GPUs.txt file and installed this into the data folder. I run the folding at home version 7.1.52 and I get nothing at all with a GTX 680. The folding slots seem to understand it now with the GTX 680 nomenclature but the work queue shows nothing. I guess I have made a mistake somewhere....but where???

[7im]

With V7, a new GPU has to be added at the GPUs.txt file level, and it has to be added at the server level. This hasn't happened yet, even though we have a ticket open to get that done.

At this point, you could try adding "client-type" set to "beta" or go back to a v6.41 client and use the "-forcegpu nvidia_fermi" and the beta team flag.

[robert2223]

I hate to be such a noobie but could you explain how to add client type to beta step by step? I have no idea where to do that.

[Jesse_V]

I hate to be such a noobie but could you explain how to add client type to beta step by step? I have no idea where to do that.

See:
viewforum.php?f=66
While I'm not entirely sure why the beta team was brought up, if you feel that adding that flag would solve your problem I would advise you to join the Beta Team. There are threads at the top of that subforum that explain how to do that. We should minimize such discussions outside of that subforum. Joining the team is encouraged if you're going to add that flag but you have to understand what that entails. We test newly-minted WUs which have just passed internal testing at Stanford. Thus, they're not as stable as the others, but we welcome volunteers to help. See that subforum for more information.

To directly answer your question, the Windows Guide illustrates how to add a flag: http://folding.stanford.edu/English/WinGuide#ntoc10

[7im]

Note that the only way to run work units on the very latest hardware is to often run the very latest work units from fah. And those are the beta work units.

If you don't want to join the beta team and do beta testing, my recommendation would be to wait for more stable work units.