Long projects

Bob Moore
Bob Moore
Joined: 5 Dec 05
Posts: 7
Credit: 8,211,394
RAC: 19,749
Topic 198121

As long as you keep sending long tasks with short deadlines, I will keep aborting them. Be less greedy and you will get more work units completed. I am sure I am not alone with this. My PC is capable, my philosophy is the limiting factor!

archae86
archae86
Joined: 6 Dec 05
Posts: 2,823
Credit: 3,298,958,266
RAC: 2,543,675

Long projects

Since it appears you regard a two week deadline as "short", your comment would be clearer if you defined what number of days deadline you regard as reasonable.

Or do you think the ratio of deadline to task execution time is a better measure? In which case what do you think reasonable for that ratio?

Lastly, BOINC can't give even somewhat reliable estimates of execution time until it has watched your work. Could it be that you pointlessly have aborted work thus depriving BOINC of the evidence it needs to provide you with reasonably accurate estimates?

ExtraTerrestrial Apes
ExtraTerrestria...
Joined: 10 Nov 04
Posts: 760
Credit: 182,460,496
RAC: 17,431

You machine should be able to

You machine should be able to complete the tasks in a few hours. A 14 days deadline seems OK for that. If the deadline was substantially longer, the database might get overloaded with too many old results which are not going to return due to some reason (i.e. the PC or BOINC installation became inactive etc.).

MrS

Scanning for our furry friends since Jan 2002

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5,210
Credit: 43,521,404,725
RAC: 44,387,428

RE: As long as you keep

Quote:
As long as you keep sending long tasks with short deadlines ...


When I first read this, I wondered if you were referring to the recent "controlled" shutdown and restart of the FGRP4 science run where indeed there were tasks with 2 day deadlines (for a couple of (mostly) good reasons :-). ). However, since you don't have any FGRP4 tasks in your list and the only aborted tasks are for the latest GW search, I guess you really are complaining about 2 week deadlines. Perhaps archae86 is correct in wondering if you are being troubled by wildly inaccurate estimates and if you would just let one finish, you might be pleasantly surprised.

Quote:
... I will keep aborting them.


That's a fairly pointless way to manage BOINC. If you don't like a particular run for whatever reason, just exclude it through preferences. At Einstein, there is not a great deal of variability between tasks for a given science run. If you don't like what you see now, there's not much point trying again later. And you really wont be able to make a proper judgement about that until you allow one to finish.

You seem to like BRP6 - you allowed one to finish and it took just 2 hours. Do you intend to get any more?

Quote:
Be less greedy and you will get more work units completed.


Oh dear!! Another one who thinks projects force hapless volunteers to be bombarded with lots of unsuitable work!!

Projects aren't "greedy". It's more a matter of the BOINC client not being suitably configured. Your BOINC client asks for work and the project does its best to meet the request. If you don't like what you get, learn to configure your client more appropriately so that you only get what you do like. Statements like this particular one (misinformed and rude) on a 'Wish List' aren't appropriate and will most likely be ignored by the Devs.

You'd be much better off if you could better state your dissatisfaction (without being insulting) and suggest reasonable changes that would make an improvement for you. You didn't even suggest what you considered to be a suitable deadline that wasn't too 'short'. You'd also need some pretty compelling reasons. The 2 week standard is pretty heavily entrenched around here.

Cheers,
Gary.

Comment viewing options

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