I've got a decently sized cache set, but i'm not getting any WU's at all... There were 45,000 ready to be sent before, now there is like 38,000. How come i'm not getting any? :(

I've got a decently sized cache set, but i'm not getting any WU's at all... There were 45,000 ready to be sent before, now there is like 38,000. How come i'm not getting any? :(

Yeah just http errors on the one My PC is trying to download. :(

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

I've got a decently sized cache set, but i'm not getting any WU's at all... There were 45,000 ready to be sent before, now there is like 38,000. How come i'm not getting any? :(

Yeah just http errors on the one My PC is trying to download. :(

I'm getting like 1 WU every 20-30 minutes now. =/

I have 4 CPU WU's running and 1 GPU (Cuda) WU running. Although the application type shows as "Not in DB" for all of them, don't know what's causing it.

I wonder why there isn't any work being sent out.

I've got a decently sized cache set, but i'm not getting any WU's at all... There were 45,000 ready to be sent before, now there is like 38,000. How come i'm not getting any? :(

Yeah just http errors on the one My PC is trying to download. :(

I'm getting like 1 WU every 20-30 minutes now. =/

I have 4 CPU WU's running and 1 GPU (Cuda) WU running. Although the application type shows as "Not in DB" for all of them, don't know what's causing it.

I wonder why there isn't any work being sent out.

The servers are sending as much work as the line can handle. See the Cricket graph pegged at 93 MBits/sec. Everybody's hosts are asking for work, the interface between the Feeder and Scheduler gets emptied very quickly so then the Scheduler says " Project has no jobs available". A few seconds later the Feeder repopulates the list, if you're lucky once in awhile your host will ask for work at the right time and get some.

When the optical fiber link becomes available, recovery from outages will probably be somewhat quicker but don't expect miracles. S@H will be restricted to only use part of that gigabit capability, and some other part of the system will probably be a limiting bottleneck anyhow.

                                                                Joe

I've got a decently sized cache set, but i'm not getting any WU's at all... There were 45,000 ready to be sent before, now there is like 38,000. How come i'm not getting any? :(

Yeah just http errors on the one My PC is trying to download. :(

I'm getting like 1 WU every 20-30 minutes now. =/

I have 4 CPU WU's running and 1 GPU (Cuda) WU running. Although the application type shows as "Not in DB" for all of them, don't know what's causing it.

I wonder why there isn't any work being sent out.

The servers are sending as much work as the line can handle. See the Cricket graph pegged at 93 MBits/sec. Everybody's hosts are asking for work, the interface between the Feeder and Scheduler gets emptied very quickly so then the Scheduler says " Project has no jobs available". A few seconds later the Feeder repopulates the list, if you're lucky once in awhile your host will ask for work at the right time and get some.

When the optical fiber link becomes available, recovery from outages will probably be somewhat quicker but don't expect miracles. S@H will be restricted to only use part of that gigabit capability, and some other part of the system will probably be a limiting bottleneck anyhow.

                                                                Joe

Thanks for the explanation. :)

I have suspended SETI for a while until the overburdened queue calms down. My log was filling with messages, all ending with project servers may be temporarily be down or Project has no jobs available. There's no point in my clogging up the wire just now with futile requests.

When things calm down (a few days?) I'll Resume SETI, get a WU, and work it.

In a larger sense, it seems that SETI is perched on a thin edge, so that any slight problem becomes a days long muddle. Even the Tuesday shut-down, known ahead as it is, causes a days long queue of volunteers' hosts requesting more WU. A queue which overwhelms the capacity of the current infrastructure.

IMNSHO, there are just too many bottlenecks lurking here.

</rant>

---

I have suspended SETI for a while until the overburdened queue calms down. My log was filling with messages, all ending with project servers may be temporarily be down or Project has no jobs available. There's no point in my clogging up the wire just now with futile requests.

When things calm down (a few days?) I'll Resume SETI, get a WU, and work it.

In a larger sense, it seems that SETI is perched on a thin edge, so that any slight problem becomes a days long muddle. Even the Tuesday shut-down, known ahead as it is, causes a days long queue of volunteers' hosts requesting more WU. A queue which overwhelms the capacity of the current infrastructure.

IMNSHO, there are just too many bottlenecks lurking here.

</rant>

I've got about 8-10 tasks waiting to start now, so i'll just stop fetching new tasks for now until it calms down a little. Seems pretty crazy at the moment, i guess it's not too bad when you have a nice cache already.

If you think about it with all the problems that s@h has had with there being no work and equipment / softwear faults you could say it is a victim of its own success... with so much work that has gone out and the fact it goes back i'm not surprised that the system gets blocked.
You know what they say " You make the best of what you'v got untill something better comes along ".