How much traffic generates Rosetta @ Home?

I'm asking myself, how much traffic Rosetta @ Home is using...

I'm looking into my internet connection stats and I've got about 100 MB per day between 6 am and 24 pm (0 am to 6 am are for free), but the only things I do is 2 hours surfing the net a day and receiving emails, which should be about 10 MB per day... I think that Rosetta is eating up my traffic volume a little bit...

I've currently 3 clients on 24/7. The last 24 hours my crunchers completed about 40 results --> about 2.5 MB per result...

What kind of numbers do the real big crunchers have out there (PY222, Bok, PCZ, ronbo54, ...)?

There have been numerous complaints over at the Rosetta@home forums due to them not compressing the data enough or sending the same files back and forth and stuff. You might find some answers over there how much traffic it generates a day. Vague answer but hopefully it will point you in the right direction.:smoking:

in the mean time you could set BOINC to only connect between 0:00 and 6:00 and set it to connect once daily.

HTH
Danny

I only Connect Twice a day, if I need more work I get it then & send in what I've done already. I keep the Network Activity Suspended the rest of the time.

Quote:

---

Originally Posted by dannychuckle

in the mean time you could set BOINC to only connect between 0:00 and 6:00 and set it to connect once daily.

HTH
Danny

---

:idea: Man, why didn't I have this idea? :D
Thanks!

OK, found out, where I could customize that, but what kind of time zone are they using? Are they using UTC? Or something other, and if yes, what?

I'm 99% sure it's the computer's local time.

HTH :thumbs:

Yes, seems to be computers local time...

Thanks for the advices! :metal:

Configured BOINC now for networktraffic between 12:00 am and 6:00 am and the ISP-traffic stats look better now :D

Need some fine tunig to not run out of work, because not all workunits have some process times and some of nodes ran out of work before 12:00 am...

I think I'll iron that out some times :cheers:

Quote:

---

Originally Posted by the-mk

Need some fine tunig to not run out of work, because not all workunits have some process times and some of nodes ran out of work before 12:00 am...

---

Go to [Network usage/Connect to network about every] and increase work cache to 10 days.

Carlos

need to watch for those large cache sizes, if we get another batch of "bad" work you could be stuck with it for awhile or be stuck manually cleaning it out. I would try 1-2 days to see if that solves your problem first.

I've configured the work cache to 2.5 days.

Now BoincView shows workbuffers of 2.5 days, 3.8 days and 7 (!!) (?) days...

So I think I wont run out of work anymore if there are no bad work units (crashing after 20 seconds)...

:hifi: