Since the Great Outage I have been running Einstein, my first foray into multiple projects.

Running 5.10.45 Boinc.

Set resource share to 100 / 100 and maintain additional work at 1 day. I assumed that since this is a Boinc setting that it would be 1 day total for all projects. However, Boinc has downloaded 1 days worth of SETI and 1 of Einstein.

Anyone shed any light on how it is supposed to work ?
Is it 1 day per project or should it be 1 day total?

Per

---

I beg to differ.

   http://dist.ist.tugraz.at/cape5/:   57423.034253  0.664618
   http://szdg.lpds.sztaki.hu/szdg/:   85775.343921  0.992770
              http://abcathome.com/:  217420.434072  2.516440
        http://www.primaboinca.com/:   77409.069431  0.895938

The middle column is the estimated CPU time (in sec), the right one is in days (86400sec), the left is obvious. Totally 4 CPU cores, access to internet is maintained by cron and allowed each 4h40m. That strange construct works something about a month (I could find exact date but I don't bother).

As you can see, over a long run boinc manages workload to be about 5day (~5.07). Thus approx. 1day worth workload per CPU core but project. And that in spite of RCN being in blinking state for all that time.

EDIT: that works since Oct14

---

I'm counting for science,
points just make me sick.

I beg to differ.

He asked about 5.10.45, not about any of the 6.x versions.

It was per, it went and became total somewhere around 6.4 (around the GPU entry), before it was turned back to per around 6.10 when the all new and shiny work schedulers were introduced.

---

It's still 6.2.14 here. I'll watch for difference as soon as I would upgrade. However, my point was it's really 1day per project just after attaching (for that project). And then the total workload approximates to 1day per CPU. But probably I miss something.

---

I'm counting for science,
points just make me sick.