Excellent news.

WCG has been pretty quiet for some time between the end of last year and first quarter of this year, the number of projects was going down. An exception with DDDT2 as a new project but in a kind of intermittent lethargic mode. Now things are speeding up. With the revival of HFCC, C4CW, CEP2, and these announcements were we see at least two projects that could come in the next months.

I imagine the workload for the tech but newer projects will attract back those who crunch until sapphire and then leave for other grids, and attract newcomers which can be inspired to crunch on these new projects. We see that C4CW is attracting chinese crunchers which otherwise would not have come.

Some of us do crunch all projects whithout preference but others are interested for personal or professional reasons to contribute to one specific project. We need to have a large spectrum of projects to attract new crunchers. It is not the other way round.

It is important to finish existing projects, but I suppose that scientific work does not need the full completion of a project, to continue working on results as they come. Before completion intermediate results can also spur new approaches, ideas, and additional tests to be performed which would lengthen the project.

Yup, that page that few of the regulars see which made it to Google too http://www.google.com/search?q=increasingly+s...cial&client=firefox-a ;>)

It's still low key though. In a few weeks the big shout goes out. Understand that WCG is prepping for 1 million results daily... cant have things buckling when the tsunami rolls in... WCG... always working hard, behind the scenes :D

If seti@home can handle the double of our crunching power and keep it supplied, then we should be able to handle that. I have nothing against searching for ET but we should one day have at least the same level of crunching power to address very immediate and important issues here on our earth.

To speak dangerously, 'bring it on'

We are actually in the process of doing some server upgrades. We should finish them off in the next 6 weeks. We are doubling the RAM for our databases (up to 64GB from 32GB) and switching from using NFS for the BOINC data filesystems to GPFS. These changes should give us some very comfortable room for upside growth. Further down the road the db servers are good to grow up to 128GB of RAM in the future and the GPFS filesystem allows us to scale the front end web servers horizontally.

In the event of what we call a 'success problem', we can pull in favors and get work done very quickly to handle the increase in load.

The biggest metric in terms of load to the infrastructure is the number of results per day. In the event of a sudden and dramatic increase in participation (say 200% growth overnight) we would increase the average runtime of each workunit. We have some projects that we can scale the size of the workunits very easily. In the event that we have sudden growth, we can lengthen those workunits and then give them more weight which slows down the result rate on the grid. This gives us breathing room while we would make server changes to handle the load.

Thank you for your responses, Kevin. Hahaha, I got you to post a response, didn’t I? Great information!

You and I go back a long time Kevin and you know I have spent considerable effort trying to get crunchers outside here to come here and give it a try. I also crunch other projects too. I don’t think anyone doubts the WCG project team’s ability to manage the hardware, software and capacity to successfully manage projects or anything unexpected. You and your team have an excellent track record.

My RNA/Orbit examples had nothing to do with WCG but examples of what happens when a project does not or cannot manage the unexpected. But these are small projects with almost no staff and even less funding.

WCG, however, does have a huge “perception problem”. Unfortunately in business, as with politics, perception IS reality, correct or incorrect. The perception outside WCG is low credits, long jail times, software incompatible with other projects and a litany of other reasons to not come here. I can use my own team as an example. We are the 11th largest team, have 3777 members and yet only 136 members choose to crunch here. Like most large teams, we go to projects as a group to crunch for that project. When we have an “Aussie Assault” here few come because their RAC drops and their world position drops. We can’t even get members to post a “G’Day” in the team thread. As with badges, people can argue to the end of time if RAC, credits, world position and the like serve any meaningful purpose in the whole scheme of things but they are reasons why people crunch. We need more people to crunch and we need more people to crunch HERE. So go kick Ari in the backside get them to fix those “perception problems”. Maybe the IBM Global Public Affairs dweebs can fix their own long-term “perception problem” of not being able to successfully market free food to starving people.

My appology for the trick, Kevin, but I got the information I wanted. Peace.



EDIT: Oh, I was serious. I hope you do get more staff. I want uplinger to get more sleep ... and send my love to Tedi