I've had no new units for a day or so now, even tho my machine is requesting, have we run out of work or is it doing a SETI? lol

16/08/2008 18:23:41|rosetta@home|Sending scheduler request: To fetch work. Requesting 47920 seconds of work, reporting 0 completed tasks
16/08/2008 18:23:46|rosetta@home|Scheduler request succeeded: got 0 new tasks

Same here, can't get any WU's, on my boxes. I hope this gets resolved quickly.

Same here, can't get any WU's, on my boxes. I hope this gets resolved quickly.

Same same here. Is there an answer to this?

---

Same here, can't get any WU's, on my boxes. I hope this gets resolved quickly.

Same same here. Is there an answer to this?

Both my computers on Rosetta, now have work. I am new to Rosetta, and have no idea what the problem was.

On my team's forum, (Overclockers.com), several new Rosetta folders commented that they were also unable to get work, as well.

Perhaps the assignment server got overloaded?

I believe it is important to keep a 4 to 5 day cache of work units for most of these BOINC projects. SETI has been infamous for not having work available, for example.

Good luck, and if you're looking for a team, we're looking for you!

There were some server problems over the weekend that caused the server to run out of work. But they seem to be resolved now.

Thanks for joining the project. Such outages are very rare, but keeping the cache of work at-the-ready assures your machine has something to do. If it works for you, something closer to a 2 day cache would be better. This is longer then most outages, and yet still gets results back to the project very quickly.

---

Rosetta Moderator: Mod.Sense

There were some server problems over the weekend that caused the server to run out of work. But they seem to be resolved now.

Thanks for joining the project. Such outages are very rare, but keeping the cache of work at-the-ready assures your machine has something to do. If it works for you, something closer to a 2 day cache would be better. This is longer then most outages, and yet still gets results back to the project very quickly.

---

Speaking of having extra work units on hand, even though I have my preferences set to 10 days worth of work, I get no more than 8 or 9 work units at a time. Hardly a days worth of processing. I lowered my preferences to 4 days thinking that this might be more realistic, but still no additional work units. What's up? Anybody know?

---

Your computers are hidden, so I couldn't look and get a feel for your runtime preference, or BOINC version. But basically the estimated runtime in BOINC has to "settle in" if you happend to make any changes to your target runtime. The BOINC client is deciding how much work to request. And the processes for making that determination have been revised in the more recent client versions of BOINC.

Sometimes, when several projects are being run, BOINC looks ahead and sees that some debt to another project should be repaid during the next 7-10 days. And so it requests less Rosetta work, and hopes to order more work for other project(s). If the other project is out of work, the client records more and more debt to that project over time. Yet there is no work to crunch to pay back the time debt. And so then, only when nearly out of work, if internet connection is available, the client will pull work from projects that have work available. But still, only enough to keep your machine busy for a while... hoping to get work for the project(s) with the debt. And so the debt problem worsens.

If you have a machine that only is connected to the internet periodically, it would probably be best to limit it to one or two projects that consistently have work available. That should help the scheduler build up the cache as desired.

---

Rosetta Moderator: Mod.Sense