Two out of my four Android BETA tasks exited with "finish file present too long". The second one that errored out just finished after 17½ hours of plodding along.

Result Name                 Status     Sent Time     Due / Return Time  CPUh/Spent Claimed/Granted
BETA_OPN1_0000335_06149_1-- Error    5/7/20 00:52:49  5/10/20 15:57:02 17.46/17.55   140.1/0.0
BETA_OPN1_0000335_03335_1-- Valid    5/7/20 00:52:49   5/9/20 11:38:28 17.44/17.55   140.2/268.7
BETA_OPN1_0000335_09673_0-- Valid    5/7/20 00:52:49   5/8/20 19:51:59 17.77/17.90   143.0/257.0
BETA_OPN1_0000335_02574_1-- Error    5/7/20 00:52:49   5/8/20 09:17:49 18.98/19.11   152.6/0.0

[Generated by wcgformat]

At first I thought the error happened because of running two tasks at the same time, but this last one ran on its own in a single-task-only queue. After its finish I enabled MCM1 and SCC1 again to fill the empty cache.

I find it worrisome that this error on my Android phone has been taking place in 50% of the cases.

Thanks for the link. A really interesting read!
Still, it's remarkable that I never had this "finish file present too long" problem on my Android phone (that processed 785 tasks since January 2019 without any error) until now.

Here are some important lines from the Result Logs:

--- Result 1 (with error "finish file present too long"): ---
INFO:[11:14:23] End AutoDock...
INFO:Cpu time = 68313.380000
called boinc_finish(0)

</stderr_txt>
<message>
finish file present too long
</message>
Reported at 5/8/20 09:17:49
(UTC "INFO:[11:14:23]" = 09:14:23)

--- Result 2 (Valid): ---
INFO:[21:50:53] End AutoDock...
INFO:Cpu time = 63987.490000
called boinc_finish(0)

</stderr_txt>
Reported at 5/8/20 19:51:59
(UTC "INFO:[21:50:53]" = 19:50:53)

--- Result 3 (Valid): ---
INFO:[13:37:40] End AutoDock...
INFO:Cpu time = 62786.000000
called boinc_finish(0)

</stderr_txt>
Reported at 5/9/20 11:38:28
(UTC "INFO:[13:37:40]" = 11:37:40)

--- Result 4 (with error "finish file present too long"): ---
<message>
finish file present too long
</message>
<stderr_txt>
(snip snip snip ... snip snip snip)
INFO:[17:52:47] End AutoDock...
INFO:Cpu time = 62859.410000
called boinc_finish(0)

</stderr_txt>
Reported at 5/10/20 15:57:02
(UTC "INFO:[17:52:47]" = 15:52:47)

--- Concluded by only one available MCM1 result (Valid) at this moment: ---
[14:46:45] Cleaning up
Result.out = 20899.000000
Run complete, CPU time: 65317.210000
called boinc_finish(0)

</stderr_txt>
Reported at 5/7/20 12:46:57
(UTC "[14:46:45]" = 12:46:45)

For uplinger:

Just received 3 Beta _2 and _3 issues with app build 7.15 on Android. Since the latest iteration is 7.17, is there any point in putting 30-32 hours into each of these when we know the 7.15 code is not picture perfect?

If not, I'd rather abort them and let someone else acquire the hours.

Edit: Each of these already have a Pending Validation copy and a No Reply + Error wingman.

Good evening again,

We are releasing a new round of work units that tests changes to the build script. It is just one batch OPN1_0000034.

Thanks,
-Uplinger

There appears to be a problem with BETA_OPN1_0000335_05220, work unit ID 80998224, running on Raspberry Pi. One host has completed, but the wingman timed out and it was passed to one of my Pies (which like the other two Pies reports the OS as Raspbian GNU/Linux 10 (buster) [4.19.97+|libc 2.28 (Debian GLIBC 2.28-10+rpi1)] ). When I first spotted it, a few hours into crunching, it had plenty of time, despite the massively reduced deadline. Last night it was at 19 hours remaining with 29 hours left on the deadline. This morning the ETA is 54 hours with 18 hours on the deadline. It is 35% complete, and that figure is very slowly increasing.

Should I abort it or let it run? Would the team like any additional information?