Smaller chunks of data
19 Oct 2010 1:15:43 UTC
Topic 195394
(moderation:
It's insane to require people to run their computers at full capacity for 30+ hours per week to be part of this project (you claim this is 15% of the week, but more accurately 100% of the time a normal person spends at home). I get the sense that there is an overwhelming support for smaller chunks of data so that people can contribute without feeling forced to keep their computers roaring all night long. I am forced to uninstall Boinc because I can't keep up with the deadlines. This is a neat project that I want to contribute to, but not willing to keep my computer running all the time.
Smaller chunks of data
)
Hi. I wish to know why I had to down load 1.5 GB of data for 31 files to processes, which was over 7 hours of download time.To days date is the 19th October and I have to complete by the 2nd November or I will not be able to receive credits. This does not give me any time to process other things I do.
Boots
RE: Hi. I wish to know why
)
Think of the data we download as a loaf of bread that then gets sliced by Boinc as it needs it, this saves the project from having to have a ton of bandwidth available for us to download each slice of bread, one at a time.
And the problem with how soon the files must be returned is why alot of us have multiple pc's now, we want to crunch for more projects and sharing causes deadline issues. Welcome to the World of multiple pc crunching it can be fun and horribly expensive but extremely rewarding too!! You will also learn alot about how pc's work, especially as your friends start giving you their old pc's and you lean how to rebuild them into working crunchers! Your place will also start looking like a pc repair shop with spare parts everywhere, BUT your rac will start climbing and you can crunch for multiple projects faster than you ever thought you could!
Well, call me Mr Grumpy, but
)
Well, call me Mr Grumpy, but I, like apparently some others, only wish to spend idle cycles of my computer. And I certainly don't build computers for that purpose, but you can, of course. The idle cycles are not enough to crunch a WU in the given time. You can either change that (reducing length of WU, increasing time) or I won't crunch anything for you.
RE: Well, call me Mr
)
Just my tuppence, of course...
Deadlines are longer in SETI.
)
Deadlines are longer in SETI. As a result, many people download too many units. I have 6 WUs in pending state in SETI, none in Einstein or other projects (5 of them, including a virtual machine). My computer is a perfectly normal one, no GPU, with a 1,8 GHz Opteron CPU running Linux.
Tullio
RE: Well, call me Mr
)
A very reasonable attitude and nobody would think otherwise.
Make sure you unsubscribe from Binary Pulsar tasks as they do take significantly longer to crunch. On your host, a GC1HF task should take maybe around 10 hours or so. If your computer is on for about an hour or two per day on average, there would be more than enough idle cycles to crunch a GC1HF task within the deadline. If you share your computer across multiple projects, that would make it more difficult. The BOINC system takes advantage of the fact that the average computer owner these days, runs their computer for at least several hours or more per day. If you run less than 1 hour per day then BOINC probably isn't for you.
With all due respect, this part of your statement is unreasonable. The E@H project is designed to process massive amounts of data from a very expensive 'big science' experiment as efficiently as possible. The raw data is exactly what it is. It is preprocessed into smaller chunks that can be delivered to large numbers of volunteered computers as well as the projects own computer clusters. It's not simply a matter of getting out a meataxe and chopping each chunk into smaller bits. To make a change that lowers the time taken to crunch a single task would likely require a complete redesign of large parts of the project. It is unreasonable to expect a project to sabotage its design goals and efficiency in order to meet the requirements of a volunteer.
The reasonable approach is for volunteers to assess a range of projects to see which ones might suit their particular requirements. Then go donate your resources to those project(s) that best suit your needs. If a project changes so that it no longer meets your requirements then change to a different project. Don't fall into the trap of making statements like "change your project or else I'm leaving". Projects tend to ignore these sorts of threats.
Cheers,
Gary.
This thread is pretty much
)
This thread is pretty much handled but I just wanted to express my feelings on whats here:
A message to all the disgruntled volunteers -
I started crunching for boinc (einstien specifically) 6 or so months ago with one pc. The hard drive crashed of no fault but age itself and I repaired it, and of course looked seriously at a refurbished machine at the pc shop. It was 4x everything I had (even with 1 cpu). I am not a gamer but I am serious about stable machines even as slow my oringinal one was. Long story short I bought that other machine - now it was boinc'in. As I learned to navigate boinc better and understand stats I realized I was just a punk compared to some newbies cranking in points. I was building machines in the 90's and still had junk in the closet (literally) not fit for todays web, but wow - they can still crunch data just fine for einstien. I now have 6 out of 7 machines crunching and I'm now going back to the other to retry (it didn't like it the first time) and there is a pile of donations. Not all came out of the closet - some were pc's and or parts donated by friends.
My thinking is this:
You are VOLUNTEERING (keyword) your pc, and time to administer.
If you believe in the cause of a particular project(s), your machine(s) would be running 24/7 - I don't think a laptop should be used personally.
You don't have to have multiple machines, but if you download a WU understand the commitment YOU have volunteered/chosen to do.
I aint fast - but I am stable
By the way I installed LINUX Fedora 14 on my slowest machine (first Linux system) and it boots faster than my fastest machine - I am so proud of my newest baby.