Hi,
My laptop run a AstroPulse task for near 400 hours and now I do not get the credits for that.I have the computers basically running 24 hours a day consuming energy just for the pleasure of doing seti@home tasks. I have also tablets, mobile and other PCs doing task too.
I am not happy for that and I am going to stop all my tasks.

---

The WU in question is 1898221889 (archive.is).

_0 and _1 (you) were both assigned the task on 12 Sep 2015 at 07:48:48 utc. The deadline for AP tasks is 25 days. That meant the report deadline was 7 Oct 2015 at 07:48:48 utc. You did not report it by the deadline, so it was issued to someone else as _2, and they crunched and returned it on 09 Oct 2015 at 19:00:41 utc. You finally reported it on 10 Oct 2015 at 07:38:55 utc, after it had been validated between _0 and _2.

After validation happens, if there are no more tasks for that WU that are in a "green" status (not beyond the deadline), then the results are assimilated into the database, the files are deleted, and the WU page stays visible to us for 24 hours after validation. You reported your task within that 24-hour period, but the files had already been deleted, so there is no way to compare your completed results to the other two machines that processed theirs.

I think the important part here is to figure out why your machine reported the WU 3 days late (making it 28 days from assigned to reported), when the CPU time was 10:01:25:55, but processing time was 15:09:27:16. If the machine is running 24/7, 15 days of processing time is still less than 25 days, so what happened? Was it more than 10 days before the crunching even started? I looked in the stderr section for the task itself (archive.is) and it looks to me that processing was stopped and restarted a couple dozen times. The problem with stopping and resuming APs is that it restarts from the last checkpoint that was completed. So that means that if you make it even 99% of the distance to the next checkpoint and then stop it, the next time it restarts, it will go all the way back to the last completed checkpoint, so you end up doing all of that processing again, which I believe is why there is such a difference between CPU time and total processing time (5 days is a lot, in the grand scheme of things).

If the laptop is going to sleep frequently, it might be worth selecting the option in BOINC manager's preferences to keep tasks in memory when suspended, which will keep them from restarting from the last checkpoint when the laptop gets woken up again. If it is actually being shut down, that will not have any effect.

But it could just be that it may not be a great idea for you to crunch APs on that machine anymore, so you'll want to go into your crunching preferences for the venue that laptop is assigned-to, and uncheck "astropulse" and uncheck "receive work for other applications" to keep that machine from getting those tasks again. I know it is frustrating that you spent so much time crunching that work, just to be told you were too late to get any credit for it. I think it was just a couple of unfortunate circumstances that all lined-up together in a bad way.

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

Thanks for you good explanation.
The laptop that was running that task in fact have a lot of sleeps. The one the is running 24/7 never sleeps since a few weeks ago as I run a web server there.
I never saw an astropulse task on the server. One of the my laptops only runs astropulse and also sleeps a lot. The laptop of this issue runs mixed task set and astropulse.
I have notice that sometimes the working % starts over again on a very low level compared to where it was before.
How do I make the tasks to stay in memory on sleep or better how do I short the %, that is how do I save more frequently the number crunch so it does not lose a lot when PCs sleeps or restart or whatever?
José

---