It's unfortunate. I had to remove MCM because I'm trying to let all the projects have their chance at my computing resources. MCM hogs most of the resources to itself. You really shouldn't assign priorities like this. Last I checked, each project hosted by WCG is a worthy cause, with no exceptions. Certainly they all deserve their fair share of resources?

res1233, it not like somebody arbitrarily assigned priorities willy nilly to the projects. There's a reason to it. I think one of the reasons is the volume of work units that the scientist's servers can handle (or something like that). I'm sure that there are other reasons (probably too technical for my transistor brain to process ) Anyway, please don't be upset about priorities and have fun crunching !!!

CJSL

Crunching for a better world...

How is it unfortunate? CEP2 can have massive result file sizes and they actually get sent directly to the scientists servers and not to WCG like the rest of the projects. There are many that couldn't handle sending a 50MB or larger result file back. I could run about 80 of those at a time and return them. There are many that could not and would struggle with a single WU.

UGM also has higher bandwidth considerations. When you download a WU, it can be around 12MB and then the result can be 2.5MB or so. Some would have an issue there as well. These large WU's can be an issue for the front-end systems for WCG. if I look at the number of UGM WU's that I complete in a day; around 800 to 900. The download is around 12MB and the results are generally from .25MB to 2.5MB. Let's use .5MB, so the download the result will be 12.5MB. 12.5MB * 800 = 10000MB a day or 10GB a day. That is 300GB a month just for UGM. To put this into perspective, that averages out to at least 115Kbps of bandwidth required 24x7.

OET is usually pretty short WU's and the results sizes are not all that bad. You also have that mobile devices can crunch this, so quite afew get sent over that way which can reduce the number for other platforms.

MCM1, the WU's and the results are usually not very large. Way back when they were creeping up in the UGM territory and WCG had an issue on the front-end storage; they were running very low and a lot of WU's couldn't be accepted at that time. The researches made some tweaks to control the file sizes better and every so often there are some large results but it isn't set after set like it once was.

So there are many reasons why one project has priority over another. As some have also stated, it also deals with what the researches can handle too. No need to go full speed and be at 50% completed but the researchers are still only looking at the 1% mark while another group of researchers could be already waiting for more results for a different project. MCM1 is at the stage where they are taking results and then issuing new work based off of those results. The other projects (with the excecption of CEP2) are not at that stage. So MCM1 has higher priority for a reason. Partly the scientists but also to keep the WCG system running like it is.

Yeah, I've noticed that MCM's priority has been steadily dropping. It's like the techs don't want to run out ahead of schedule, and they're slowly leaning on the brake. Once people can't do cancer and HIV, some are gonna bail. So keeping them in the mix for as long as possible may maximize the number of processor years contributed.