|
1)
Message boards :
Number crunching :
Upload fails
(Message 4239)
Posted 8 May 2020 by Leadeater Post: This is not an excuse, I'm just writing this to explain what happens. Thanks for keeping us up to date on the situation. Anything we can do to help the situation? Is it just a situation of a large bulk of work trying to be returned or is the sustained load currently just too high? |
|
2)
Message boards :
Number crunching :
Upload fails
(Message 4237)
Posted 8 May 2020 by Leadeater Post: Disc full again? ULX jobs can be as large as 20MB, there are 34008 of those in progress, average that out so 340GB and that's not taking in to account the hourly rate those are released. Then you have to account for storage space required for other process to get jobs sent and what happens once received. BHspin might be a lot smaller but there's high hundreds of thousands of those, no idea how many released per hour. Basically a 120GB won't cut it, I would expect a few TB is required at a minimum, just for the active data. Who knows how much is kept long term. |
|
3)
Message boards :
Number crunching :
Upload fails
(Message 4235)
Posted 8 May 2020 by Leadeater Post: This is very close minded. I began running BOINC after seeing a post on my forums looking for more members for a Formula Boinc sprint challenge. With teams looking for the top spot during the pent, many start recruiting members within their forums, websites and circles of friends. Some do the pent and then never come back, others see the interest and continue. Well said. On the project admin side it's not like everyone is flush with funding and can put a lot of resources in to infrastructure and gauging the expected load demand isn't always that straight forward. Just have to be patient and understanding, it's not worth getting angry over. Even if all my work ends up not counting towards the Pent I'm still glad I did it and I'll still make sure it gets uploaded. |
(English)
(Polski)