This morning I thought that I would try to crunch a few more WU's for Rosetta -

So I opened the flood gates on 40+ crunchers - only to find my internet line bogged down due to the very large WU's. Most of these WU's are over 3 meg in size -

3 MEG ??????? This is worse that SIMAP !

I am forced to abort the downloading on over half of my crunchers due to the extreem size of these WU's

---

SETI.USA

This morning I thought that I would try to crunch a few more WU's for Rosetta -

So I opened the flood gates on 40+ crunchers - only to find my internet line bogged down due to the very large WU's. Most of these WU's are over 3 meg in size -

3 MEG ??????? This is worse that SIMAP !

I am forced to abort the downloading on over half of my crunchers due to the extreem size of these WU's

They have crept up a little, but these are a smaller than some of the units we got. I think any usint starting boinc_ has had a file size reduction. I don't know about the newer ones. I is something I pressed on a lot about, running 4 computers over dial-up was a pain in the backside. the boinc_ 1Mb units where a god send for that. Though the dial-up has recently gone, the rest still on dial-up have had to leave rosetta now :-(

Set your workunit length to 24hrs and stagger the launching of the clients... this will reduce overall bandwidth. (the workunit size stays the same so you save about 4 times over the default setting.)

---

Team mauisun.org

Hi Kevin

If you want to reduce the bandwidth I think setting up one of your machines as a proxy would help you as the files would only need to be downloaded once each no matter how many clients were using them. If that'd be useful to you then I'm sure someone here can post details.

Danny

---

This morning I thought that I would try to crunch a few more WU's for Rosetta -

So I opened the flood gates on 40+ crunchers - only to find my internet line bogged down due to the very large WU's. Most of these WU's are over 3 meg in size -

3 MEG ??????? This is worse that SIMAP !

Well... a quick test reveals a SIMAP HMMER-wu is 5.69 MB downloaded, and while not sure on actual run-time it claims 2.5h before starting on an old 1.4GHz-cpu. This means 9.6 wu/day, each of 5.69 MB, for a total of... 54.624 MB/day downloaded.

Rosetta@home on the other hand, here user can choose to set Target CPU run time to 24h, meaning roughly 3 MB/day downloaded.

3 MB/day << 54 MB/day...


The other type of Simap-wu is smaller, but is also using less time to crunch, so will still use more MB/day than Rosetta@home.

It's only if you've configured Rosetta@home to 1h-run-time that Rosetta will download more MB/day than SIMAP.

---

"I make so many mistakes. But then just think of all the mistakes I don't make, although I might."

3 MEG ??????? This is worse that SIMAP !

Not so, I joined SIMAP for a month until I noticed I was sending/recieving over 3GB of data per day! (Holy smokes!) Now that I'm back here it's down to a more manageable 300 - 400 MB's.

If you want to reduce the bandwidth I think setting up one of your machines as a proxy would help you as the files would only need to be downloaded once each no matter how many clients were using them.

A good idea, although I'm not sure that will work for the WU's. As was said earlier - another trick is to increase your WU time so it runs them longer. If you use 12 hour WU's you should see a lot fewer 3MB WU's than the standard 3 hour setting.

Welcome back Kevin!

---

Team Starfire World BOINC