Like many other people i've been getting "aborted by project" wu's for about a week, but in the last 36 hours they seemed to have stopped. Or so I thought, as the Boinc message log showed no sign of any. but it turns out I've had over 90 aborted by project without any mention of it in the log,they even report,and dissapear imediatly.I only found this out because they record in boincview completed log.I double checked with my computer results page and sure enough they're they were.

---

m4rtyn
**************************** ***************************

Like many other people i've been getting "aborted by project" wu's for about a week, but in the last 36 hours they seemed to have stopped. Or so I thought, as the Boinc message log showed no sign of any. but it turns out I've had over 90 aborted by project without any mention of it in the log,they even report,and dissapear imediatly.I only found this out because they record in boincview completed log.I double checked with my computer results page and sure enough they're they were.

I suspect it's going to be difficult to tell what the issue might be because it appears your computers are hidden, therefore no one can see your results and the messages.

---

Like many other people i've been getting "aborted by project" wu's for about a week, but in the last 36 hours they seemed to have stopped. Or so I thought, as the Boinc message log showed no sign of any. but it turns out I've had over 90 aborted by project without any mention of it in the log,they even report,and dissapear imediatly.I only found this out because they record in boincview completed log.I double checked with my computer results page and sure enough they're they were.

I suspect it's going to be difficult to tell what the issue might be because it appears your computers are hidden, therefore no one can see your results and the messages.

Unfortunatly ther's not much help their anyway,All 4 computers atatched to seti show the same -221 exit status(The project you are attached to aborted this result.) as with the earlier aborts that came with various entries in most peoples message logs. Also Wu's mostly show two credited rusults from other uses and no credit on mine, and all were well within the deadline. Very strange!


PS. Computers are not set as hidden in my prefs, Have to look into that.

---

m4rtyn
**************************** ***************************

Just looked, and your computers are out in the open.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Just looked, and your computers are out in the open.

Thanks for letting me know, I was a bit puzzeled by that.

---

m4rtyn
**************************** ***************************

Unfortunatly ther's not much help their anyway,All 4 computers atatched to seti show the same -221 exit status(The project you are attached to aborted this result.) as with the earlier aborts that came with various entries in most peoples message logs. Also Wu's mostly show two credited rusults from other uses and no credit on mine, and all were well within the deadline. Very strange!


PS. Computers are not set as hidden in my prefs, Have to look into that.

Not showing up in the log is kind of weird, but the deadline or how much 'margin' your host has in getting it done on time is irrelevent when it comes to a 221 useless result abort.

The only two factors which count are:

1.) Has the WU validated successfully?

2.) Is the result still unstarted on the host?

If both are true, the result will 221. Anything else should still run to completion.

Alinator

Unfortunatly ther's not much help their anyway,All 4 computers atatched to seti show the same -221 exit status(The project you are attached to aborted this result.) as with the earlier aborts that came with various entries in most peoples message logs. Also Wu's mostly show two credited rusults from other uses and no credit on mine.


PS. Computers are not set as hidden in my prefs, Have to look into that.

Hmm.. I looked at a couple of your aborted results. It appears that they aborted before you started running them (0.00 sec cpu time). It also appears that the quorum was met (2 results validated) well before they were aborted on you as well. It looks like this may be the new intended behavior of the system. They didn't 'need' your result, so they aborted it before you started crunching on it. You lost zero crunching time, and they get to clear that entry out of the database that much quicker. Now, from what I understand, if you had already started that result (non-zero cpu time) when you contacted the scheduler, your result would not have been aborted and you would have been able to report that result when completed for credit (if, that is, it passed validation).

To repeat, you didn't lose ANY crunching time, because you had not yet started processing those results. Your BOINC client should just download new ones to replace those aborted ones, if the project has more work available. This will help ensure that the work everyone does is actually *useful*, and not just 'busy-work'.

---

https://youtu.be/iY57ErBkFFE

#Texit

Don't blame me, I voted for Johnson(L) in 2016.

Truth is dangerous... especially when it challenges those in power.

To repeat, you didn't lose ANY crunching time, because you had not yet started processing those results. Your BOINC client should just download new ones to replace those aborted ones, if the project has more work available. This will help ensure that the work everyone does is actually *useful*, and not just 'busy-work'.

Two concerns:

1) Running out of 'busy-work' will likely lead to complete outage of work sometime real soon based on what has been said, unless more old data is split.

2) Does this impact the daily quota? I haven't read enough (due to slowness and downtime) to see if it has been stated definitively yes or no.

To repeat, you didn't lose ANY crunching time, because you had not yet started processing those results. Your BOINC client should just download new ones to replace those aborted ones, if the project has more work available. This will help ensure that the work everyone does is actually *useful*, and not just 'busy-work'.

Two concerns:

1) Running out of 'busy-work' will likely lead to complete outage of work sometime real soon based on what has been said, unless more old data is split.

2) Does this impact the daily quota? I haven't read enough (due to slowness and downtime) to see if it has been stated definitively yes or no.

First off, the old data tapes we're currently running is not the same ones that were reissued adinfinatum on Classic. They are ones which never got fully processed and assimilated into the MSD on Classic and/or the 'low-res' original BOINC apps.

Secondly, yes. Currently a 221 auto abort counts the same as any other abort or client error, -1 in the qutoa column.

Alinator

To repeat, you didn't lose ANY crunching time, because you had not yet started processing those results. Your BOINC client should just download new ones to replace those aborted ones, if the project has more work available. This will help ensure that the work everyone does is actually *useful*, and not just 'busy-work'.

Two concerns:

1) Running out of 'busy-work' will likely lead to complete outage of work sometime real soon based on what has been said, unless more old data is split.

2) Does this impact the daily quota? I haven't read enough (due to slowness and downtime) to see if it has been stated definitively yes or no.

First off, the old data tapes we're currently running is not the same ones that were reissued adinfinatum on Classic. They are ones which never got fully processed and assimilated into the MSD on Classic and/or the 'low-res' original BOINC apps.

Secondly, yes. Currently a 221 auto abort counts the same as any other abort or client error, -1 in the qutoa column.

Alinator

I think I had understood that it wasn't a reissue "Classic-style", but thanks for pointing that out. I would still think that this process would cause new data to get picked up even faster than before... Of course, this could be the desired impact; to run dry as an attempt to "encourage" some hosts/users off towards other projects. It is highly interesting to note the front page has an article about all BOINC projects available...

Coincidence? Maybe........

Secondly, yes. Currently a 221 auto abort counts the same as any other abort or client error, -1 in the qutoa column.

just to confuse the issue a bit, despite around 90 to 100 -221 aborts in the last 36 hours and alot more over the last week, none of my computers has had a daily quota reduction recently

---

m4rtyn
**************************** ***************************

Probably you just didn't see it, because every successful valid result doubles the quota again. 7 success's takes you from 1 back 100 again.

Alinator

Secondly, yes. Currently a 221 auto abort counts the same as any other abort or client error, -1 in the qutoa column.

just to confuse the issue a bit, despite around 90 to 100 -221 aborts in the last 36 hours and alot more over the last week, none of my computers has had a daily quota reduction recently

As I said, I can't find a "definitive" yes or no... ;-)

One possibility though is that it did cause the reduction, but the return of good results is offsetting the reduction...

One possibility though is that it did cause the reduction, but the return of good results is offsetting the reduction...

BINGO!

I think they recently increased the maximum quota possible from 4 CPUs worth to 8 CPUs worth of units (4x100 to 8x100 maximum possible). The only way the quota is gonna bite you due to these project-aborted results is if you have a huge number of CPUs in that host (16 or more), the host's CPUs are insanely fast (a high clockspeed Intel 'Core 2' series, for example), and you run an insanely high cache. For most participants, with 1 or 2 cpu systems, this is a non-issue. Even if you get 50 work units canceled by the project in one fell swoop (dropping your quote from 100 to 50), the next return of a single valid result will double that again to 100.

---

https://youtu.be/iY57ErBkFFE

#Texit

Don't blame me, I voted for Johnson(L) in 2016.

Truth is dangerous... especially when it challenges those in power.

As I said, I can't find a "definitive" yes or no... ;-)

One possibility though is that it did cause the reduction, but the return of good results is offsetting the reduction...

Zombie confirmed it happened when the feature was first turn on a few days ago.

Alinator

One possibility though is that it did cause the reduction, but the return of good results is offsetting the reduction...

BINGO!

I prefer cards, specifically Spades... ;-)

UNO is fun though...and it certainly is applicable to this situation...

Skip....Draw Two...Reverse (abort)...Draw Four (Red...signifying more aborts)... UNO!

One possibility though is that it did cause the reduction, but the return of good results is offsetting the reduction...

BINGO!

I prefer cards, specifically Spades... ;-)

UNO is fun though...and it certainly is applicable to this situation...

Skip....Draw Two...Reverse (abort)...Draw Four (Red...signifying more aborts)... UNO!

LOL....

Throw Mille Bornes in there too. 221, you just got a 'Panne D'Essence' thrown on you, or maybe a 'Creve'. :-)

Alinator

I think they recently increased the maximum quota possible from 4 CPUs worth to 8 CPUs worth of units (4x100 to 8x100 maximum possible).

I don't think the max has been raised above 400. I still hit the daily limit for several days after each outage. Same as before when it was 400 for sure.

---

Dublin, California
Team: SETI.USA

I think they recently increased the maximum quota possible from 4 CPUs worth to 8 CPUs worth of units (4x100 to 8x100 maximum possible).

I don't think the max has been raised above 400. I still hit the daily limit for several days after each outage. Same as before when it was 400 for sure.

While I'm still not in favor of raising that cap, particularly since work now is in scarce supply, I would like to say that the project-side aborts IMO shouldn't drop your quota limit. So long as the good result reports are able to offset the negative impact, then I personally don't have a problem with what's being done.

I think they recently increased the maximum quota possible from 4 CPUs worth to 8 CPUs worth of units (4x100 to 8x100 maximum possible).

I don't think the max has been raised above 400. I still hit the daily limit for several days after each outage. Same as before when it was 400 for sure.

From the BOINC changelog:

5486 David 29 May 2007
5487 - client: add cmdline options --exit_before_start and --exit_after_finish.
5488 This cause the core client to exit immediately before or after
5489 running a job,
5490 letting you examine the contents of the slot directory.
5491 - scheduler: changed max # of CPUs used in daily_result_quota
5492 limit from 4 to 8, and make it a compile-time parameter

I don't know if the version of the scheduler code running on the main project includes this yet. But it was done a couple of weeks ago, so there is a chance it does.