Now that you mention checkpointing problem .....
I remember that I have detected it too.
Might have the same cause!
In my case on Win7 x64 prof.EN , on latest betaver. boinc .29+ :
Project: Asteroids@home
Symptom: every time I restarted my PC (I closed Boinc manually way before
the shutdown), I noticed that ALL workunits of Asteroids started from
zero percent. I noticed this in 100 percent of restarts , at least 3 times
now in the last 2 weeks.
So, I decided to report the problem.
Please, can someone confirm this? I mean not NOW but ....... once the
Asteroids server is back in full-operation. (It has had run out of disk
space, recently). :)
I will definitely try to downgrade and test it on previous Boinc version
*soon* to see if the behavior was correct back there.
' will let ya know...
Thank You
Filip
Dne 26. 5. 2016 6:12 odpoledne napsal uživatel "Eric Driver" <
[email protected]>:
I've noticed that sometimes when the client starts a fresh work unit, it
will
read the check point file of a previous work unit. Not sure how/why this is
happening, but this is obviously not a good thing. One way I can fix this
is to
use a checkpoint filename unique to the WU (up until now all WUs use the
same
generic name). I'd like to key off of the work unit dat filename, but it
is not
immediately apparent how to get my hands on that information from inside my
main
app. Can someone point me to an easy way to do this, or possibly another
solution to this problem.
Thanks!
Eric
_______________________________________________
boinc_dev mailing list
[email protected]
http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev
To unsubscribe, visit the above URL and
(near bottom of page) enter your email address.
_______________________________________________
boinc_dev mailing list
[email protected]
http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev
To unsubscribe, visit the above URL and
(near bottom of page) enter your email address.
