Aborting a (lot of) WUs - without 'consequences'?

Denis Puhar, dr. med.
Denis Puhar, dr...
Joined: 5 Nov 09
Posts: 36
Credit: 7006583
RAC: 0
Topic 196637

Hi!

Just by chance I spotted this - at least for me - rather unusual practice:

I know, that sometimes we must all ABORT a WU (or even several ones) for various software or hardware reasons (new components, drivers,...).

But I've NEVER came across a fact that my 'partner' in doing an 'GPU Arecibo' task ABORTED almost a couple of hundred WUs for no apparent (!?) reason.

And this applies not only for WUs he received let us say one day all of them together, but ALSO there were aborted WUs, he received (I assume together) the next day - ALL of them!

1. Such 'behaviour' slows down the process of VALIDATION considerably (with consequences on all of volunteers);

2. I can't stop asking myself, if there is a more specific reason for someone to 'engage' in such a 'behaviour' (albeit I have no clue what that would be)!?

And I'm pointing out again:

We ALL have sometimes for such or another reason abort WUs and I do not have a problem with that.

But when the figure is so high and the the 'owner' behind the computer is John Doe - 'Anonymous', I decided just to ask, if such 'methods' are simply 'overseen' without any question being asked?

Denis

“A little knowledge is a dangerous thing. So is a lot.” - Albert EINSTEIN

ggesmundo
ggesmundo
Joined: 3 Jun 12
Posts: 31
Credit: 18699116
RAC: 0

Aborting a (lot of) WUs - without 'consequences'?

I can't answer for others but, I recently attached to Boincstats BAM and without any warning all of my current tasks were aborted. Certainly not what I had intended, just wanted to be able to adjust things while on the road. Apologies to all wingmen affected but, sometime crap happens.

archae86
archae86
Joined: 6 Dec 05
Posts: 3157
Credit: 7213554931
RAC: 971868

I don't know what

I don't know what consequences would satisfy Denis, but as I recall there is a consequence in that aborting work knocks down the allowed daily work fetch--which then recovers as one reports normally completed work.

The extra network traffic expended to deliver the aborted work to the original user seems to me the only material resource bearing on project productivity cost wasted. Any policing measure considered would need to be a net improvement--considering cost to agree on it, and to develop it, the continuing cost of operating it, and net effect on users.

I don't think this problem is crying out for a revised solution.

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3522
Credit: 713758763
RAC: 902047

Yup, I guess one should not

Yup, I guess one should not jump to conclusions. Apart from technical mishaps like the one described above, there are even GOOD reasons to abort jobs. If you find that your PC has received more jobs than it can possibly handle (e.g. you want to go on vacation, or you want to change your share between projects, or whatever...) it's actually the FRIENDLY thing to do to abort the tasks, and it will speed up validation because the only alternative (let the tasks timeout after two weeks) is definitely worse. In this sense, I don't want to discourage aborting tasks generally.

There are some projects where some volunteers do selective aborting of tasks to 'cherry-pick' tasks with a high credit to runtime ratio, but this is not an issue at E@H currently because

a) all tasks of a particular search have pretty much identical runtime on the same hardware and
b) you can de-select whole searches if you prefer tasks from certain seaches over others (for whatever reason).

Cheers
HB

Horacio
Horacio
Joined: 3 Oct 11
Posts: 205
Credit: 80557243
RAC: 0

Sometime ago I had to abort

Sometime ago I had to abort hundreds of WUs due to a failure in a GPU that dragged the DCF to a very very low value and then BOINC thought that I can process much more than the host was really able.

Of course, Ive fixed first the GPU thing, then waited for the DCF to rise at a reassonable value and then Ive aborted the excesive tasks, this way BOINC don't asked for more task just after aborting.

In those cases is better to abort than to wait until they expire, as the aborted task will be resent to another hosts sooner than if they were left there until the deadline.

The way in BOINC handles the "punishments" when a user abort tasks is by decreasing the number of tasks they can get per day, but this limits rise pretty fast so is not a very effective meassure to discourage users about aborting tasks...

By the way, if the host does not belongs to an annymous user, you can try to send a Private Message to the user offering him to visit the forum in case is having some issue... (of course if the user dont read the forums and has their account set to not send a mail on PMs then he will never read you message)

EDIT: Im very slow writting... LOL

Denis Puhar, dr. med.
Denis Puhar, dr...
Joined: 5 Nov 09
Posts: 36
Credit: 7006583
RAC: 0

Sorry, not to report for some

Sorry, not to report for some quite time.

But I guess this was in the end a good thing after all. Why? To read all of the answers and reply to all of them at once.

I think my message was a bit misunderstood.

1. I did NEVER use a word 'PUNISHMENT' when such things happen. And I did not 'point a finger' into anyone at any time.

2. With 'consequences' I simply implied a 'scenario' (which at my knowledge did NOT occur, but nevertheless at seeing what I described, it made me simply curious and that was the reason for a question, not for proposing what SHOULD be done) in which a person XY receives hundreds of WUs, aborts them for no apparent reason, gets a hundred NEW WUs and does the same again.

Certainly we agree, that if:

A) Such things happen repeatedly for no apparent reason;
B) If let us say such XY person is not ONE individual, but several ones ore even more;

That this will have some negative consequences. But what or even if anything should be done in such a case is not my business, nor I have the necessary skill even to propose a course of action (or if it is needed).

The SOLE reason in writing my message was CURIOSITY, because I've NEVER seen such a thing happen on such a LARGE scale and that REPEATEDLY.

I also pointed out TWICE, that we all experience 'scenarios', when for such or another reason we 'must' (or that is the best course of action) abort WUs.

And the example I've been given in the first reply is certainly a completely VALID reason and there was also absolutely no reason to apologise to me for something, that happened also to me (not to mention various other reasons I sometimes have, not only in E@H, but also in other projects to do so).

But as I understand threads are here for asking questions?

Aren't they?

And this is the ONLY thing I've done (although my choice of words was maybe poorly chosen - English is not my native language), there was not even a SLIGHT intent to make objections or even accusations on anyone, because I know too little, about how the whole process even works.

Anyway, thank you for the answers from all and apologies, if I was misunderstood in the intent of why my first message was even written.

Regards.

Denis

“A little knowledge is a dangerous thing. So is a lot.” - Albert EINSTEIN

ggesmundo
ggesmundo
Joined: 3 Jun 12
Posts: 31
Credit: 18699116
RAC: 0

I for one am glad you asked,

I for one am glad you asked, discussions like these help us all gain a better understanding of how things work. I have not found a better source of information than the forums.

Denis Puhar, dr. med.
Denis Puhar, dr...
Joined: 5 Nov 09
Posts: 36
Credit: 7006583
RAC: 0

Thank You, I really

Thank You, I really appreciate your opinion.

I think the same way and even think that the importance of such forums is more often than not under-appreciated.

Why? At least in my experience from other project(s) forum(s) - no need to go into detail - but at one point, instead of being an example of what you perfectly described in ONE sentence, it literately turned into a 'tool' of verbal abuse and a place of 'double standards' (where some individuals 'exploited' their status and knowledge in very disappointing way to say at least).

In spite of being a volunteer even before BOINC came into existence (for SETI - if I'm not mistaken - more than for 10 years), one SINGLE such event almost brought me to the state, when I was willing to leave EVERYTHING in connection with BOINC behind.

I was even surprised at my own (over)reaction (being an ER physician and used to pretty much everything...).

But EXACTLY that was the source of my reaction I suddenly realised.

It was a kind of a 'deja vu'. In a way, where some people exploit their 'privileged' positions to reach their goals (whatever they may be) or use it as an 'excuse' to have the right to insult someone without consequences. And the people in charge, who could and should have done something to prevent such things, rather look the other way in the 'greater interest'. Nothing new. Even common in 'my world'.

But to experience something similar (at least in a way that reminded me of such 'practices' and where suddenly 'The Code of conduct' was nothing more than a formality with no practical use) at a place, which should be a definition of cooperation at expense of no one was a little bit too much for me.

And yes. That was my problem and I may have overreacted. But no matter from which perspective someone looks at such 'events', it certainly does not help to gain volunteers, rather the opposite.

I excuse myself for such a long reply. It does not even belong in this thread (a personal message would be better), but I think, no harm was done.

All I wanted, was to express the importance of what you summarised excellent in one sentence and show with a personal experience (by pointing at no individual or project explicitly) how really important such forums can be in a positive sense, but also 'dangerous', if not used properly, with some regard to people, who aren't scientists, don't have proper knowledge, do sometimes ask 'stupid' questions and make other mistakes, but more than enough, they overcome those 'shortcomings' with their willingness to help, enthusiasm and curiosity with one intention only:

To help in a common effort, that is much bigger than every single one of us is!

Regards.

Denis

“A little knowledge is a dangerous thing. So is a lot.” - Albert EINSTEIN

David S
David S
Joined: 6 Dec 05
Posts: 2473
Credit: 22936222
RAC: 0

I don't read the E@H forums

I don't read the E@H forums often (usually, like now, I'm looking for info about a major outage at Seti), but this thread caught my eye.

A week or two ago, my i7 computer aborted 99 E@H units all on its own. I didn't do it. I was very surprised to check my account one morning and find that I had aborted a bunch of work. I could only guess that it didn't think it could do them by deadline or even that the deadline had passed already. (The same day, I also had one unit marked invalid because it was returned too late.)

So in the case of the host you're inquiring about, it may be that it did it by itself and the owner is innocent, maybe doesn't even know it happened.

David

Miserable old git
Patiently waiting for the asteroid with my name on it.

juan BFP
juan BFP
Joined: 18 Nov 11
Posts: 839
Credit: 421443712
RAC: 0

RE: I'm looking for info

Quote:
I'm looking for info about a major outage at Seti)


Posted in the front page of SETI:

Unscheduled Power Outage
Power to the north east side of the Berkeley campus and the Space Science Lab was interruped this morning at about 16:30 UTC until about 17:30. The cause of the outage has not yet been determined. We're working to get our databased back in working order. Current estimate is that we'll be back online on Tuesday, Dec 6. 30 Nov 2012 | 4:42:45 UTC

So expect a looong stay here... so is better to free some cores in you CPU to help the GPU E@H WU running faster on this days...

lHj2ixL.jpg

 

Claggy
Claggy
Joined: 29 Dec 06
Posts: 560
Credit: 2699403
RAC: 0

RE: A week or two ago, my

Quote:
A week or two ago, my i7 computer aborted 99 E@H units all on its own. I didn't do it. I was very surprised to check my account one morning and find that I had aborted a bunch of work. I could only guess that it didn't think it could do them by deadline or even that the deadline had passed already. (The same day, I also had one unit marked invalid because it was returned too late.)


I did have a look at your host when you posted this at Seti, Basically Boinc aborted all that work because the deadline had passed (two weeks), and because Einstein is running older server software it can't tell the difference beteeen Boinc aborting the work and you aborting the work.

Error tasks for computer 4124638

Claggy

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.