Quote from: rastiks on April 19, 2018, 03:00:07 PM

Quote from: 616westwarmoth on April 18, 2018, 06:16:44 PM

Quote from: jaapgvk on April 17, 2018, 10:05:17 AM

Quote from: 616westwarmoth on April 16, 2018, 10:00:25 PM

Quote from: capulo on April 16, 2018, 09:32:23 PM

is there any problem with rosetta tasks again? all my computers have insufficient tasks amount Sad

I've not noticed it on my end and while the available tasks are a bit below average on the server, https://boinc.bakerlab.org/rosetta/server_status.php shows 2026 unsent Rosetta tasks and 13234 Rosetta Mini tasks so you shouldn't be having an issue.

My test machines are supposed to arrive tomorrow and I'll split them evenly between WCG and R@h so in a few days should have a good baseline for what is to be expected between the two for determining a fair RAC equivalent. But in the meantime, you might consider throwing some power at WCG as a hedge. The R@h project we're tending to be in the 15% work performed range (we solved 13% of the work units overall yesterday), whereas if we did a comparable amount of work at WCG (roughly 5M RAC) we'd be around 3-5% (although right now we're less than 1%). That is to say the WCG project can handle the influx much better than R@h can.

Sadly, my instances were terminated before I could get any accurate data. I don't have any similar machines at my disposal anymore and sadly I cannot afford to start again at this point.

So far I've got 2 smaller machines that have been up for a week and are showing mid 2300 under host average credit (in BOINC), and the two identical machines to those were showing 3800 RAC.

I did fire up four identical heavier computers last night ,two on R@h and two on WCG, it will take until the end of the weekend to begin to get meaningful data from those.

Well, I can share some numbers too. I have a 12/24 core Xeon X5650 server which was running Rosetta tasks for ~30 days and during that period its maximum RAC climbed to 9612. Then I moved it to RAH:WCG in 50:50 ratio and about a week ago, I moved it to 100% WCG. The current RAC is:

$ boinccmd --get_project_status | grep host_expavg
host_expavg_credit: 2619.250212
host_expavg_credit: 6877.862939

2619 is for RAH, 6877 is for WCG. With * 1.5 factor for WCG, this particular server now produces RAC = 1.5 * 6877 + 2619 = 12934 copared to the 9612 when it was running Rosetta only (and I expect it to be even more as RAH decreases and WCG increases, I can update later). For this server, it seems that RAH RAC = WCG RAC.

don't forgot that there was insufficient tasks issue with rosetta in last month, most problably this affected your rac. all my x5650 was able to do rosetta rac over 12000
and maybe graphs for gaining/loosing rac are not same for rosetta and wcg...
we must run same machines from zero and then compare them