Problem With SMP

im getting a new error with Windows SMP Client everytime i run it im getting this



[18:45:54] Completed 40000 out of 1000000 steps (4 percent)
[19:08:17] Writing local files
[19:08:18] Completed 50000 out of 1000000 steps (5 percent)
[19:30:42] Writing local files
[19:30:43] Completed 60000 out of 1000000 steps (6 percent)
[19:53:10] Writing local files
[19:53:10] Completed 70000 out of 1000000 steps (7 percent)
[20:16:34] Writing local files
[20:16:34] Completed 80000 out of 1000000 steps (8 percent)
[20:41:52] Writing local files
[20:41:52] Completed 90000 out of 1000000 steps (9 percent)
[21:06:42] Writing local files
[21:06:42] Completed 100000 out of 1000000 steps (10 percent)
[21:31:14] Writing local files
[21:31:14] Completed 110000 out of 1000000 steps (11 percent)
[21:55:05] Writing local files
[21:55:05] Completed 120000 out of 1000000 steps (12 percent)
[22:18:32] Writing local files
[22:18:32] Completed 130000 out of 1000000 steps (13 percent)
[22:26:34] Gromacs cannot continue further.
[22:26:34] Going to send back what have done.
[22:26:34] logfile size: 82886
[22:26:34] - Writing 83422 bytes of core data to disk...
[22:26:34] ... Done.

What WU?

its 2652

Hey Roger ... sorry i dropped out for a few days last week due to RL stuff.

I am guessing from your log snippet that the wu you are running is either a 2652 smp wu or not an smp wu .... did you install the v 6 win SMP beta client?

If you are running the v6 client, it is not yet supporting smp, at least not officially.
For SMP on win at the moment I would suggest running the current 5.91 win smp beta cleint.

Also, when you post a failed wu, it is most helpful if you include the whole section from the startup of the client. Surround it with code tags like this:


just quote my post to see the code tags.

XRogerX said:

its 2652

Click to expand...

A few 2652's have had problems ... it should restart the wu, and probably fail at the same point again.

After 2 failures at the same % complete on the same wu (ie 2652 with same run,clone,gen #'s) you can save time by stopping the client, deleting the wu (by deleting the work dir and queue.dat file). and restarting the client. You may have to repeat this till the client gets a different wu.

Alternatively, let it just recrunch it up to 4 times then it should get a new wu by itself although sometimes these malformed wu's will hang and need to be deleted as described above.

pscout said:

A few 2652's have had problems ... it should restart the wu, and probably fail at the same point again.

After 2 failures at the same % complete on the same wu (ie 2652 with same run,clone,gen #'s) you can save time by stopping the client, deleting the wu (by deleting the work dir and queue.dat file). and restarting the client. You may have to repeat this till the client gets a different wu.

Alternatively, let it just recrunch it up to 4 times then it should get a new wu by itself although sometimes these malformed wu's will hang and need to be deleted as described above.

Click to expand...

Well pscout i did what you said with the whole deleteing part and its wroking so far ,the run part was always at 0, i deleted it like four times and boom i got one with a run or 30 this is a good sign ,im up to 20% now this is the longest it has been running without a problem i see this WU has some nice points to it ,well lets see what happens when it finishes ,and you all know im going to post back if theres a problem lol, theres a shocker lol

congrats but I just had one stop at 70+% also...

learning smp myself...