recent sever status shows 200000+ WUs read to send.
Why is that?
boinc ver. 6.12.34

recent sever status shows 200000+ WUs read to send.
Why is that?
boinc ver. 6.12.34

The feeder only holds 100 Wu's at once, when your host asked for work it was empty, a moment later it would have refilled.

Claggy

I have done this 10 times now.
I am only getting GPU units (when i abort some), no cpu tasks.

I am only getting GPU units (when i abort some), no cpu tasks.

I See, you now have a Max tasks per day of One for 'SETI@home Enhanced (anonymous platform, ATI GPU)'

Claggy

Between "Project has no tasks available" & "HTTP service unavailable" I can't seem to keep any of my rigs stocked up on GPU task. Would be nice to get them both full before the Tue outage, but there currently both at zero GPU task.

Any ideas?

---

My Computers:
â–ˆ Blue Offline
â–ˆ Green Offline
â–ˆ Red Offline

I've found that when I request new tasks, about 85% of the time I get "project has no tasks available". The 15% of the time I do get tasks, more often than not I get a flood (30+ tasks). Are there rules in place that govern how many of those 100 tasks 1 person can gobble up? I'd be satisfied knowing that there was a greater chance of getting a few WUs each time I requested, than taking the gamble and coming up empty most times.

I don't know about "rules", but I've had more than 30 delivered in one hit on a few occasions in the last few weeks.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

I have done this 10 times now.
I am only getting GPU units (when i abort some), no cpu tasks.

Why are you aborting tasks, and then complaining that you aren't getting any?

---

true just use the rescheduler to move the work around on that machine

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

I have done this 10 times now.
I am only getting GPU units (when i abort some), no cpu tasks.

Why are you aborting tasks, and then complaining that you aren't getting any?

I can't speak for him, but the one time I did get some listed under the transfers tab, they just sat there and wouldn't download. Over and over again I hit the retry button and it does absolutely nothing. I thought cancelling the download would fix it, but it didn't. Now it gets absolutely nothing. I switched over to a different project for now, but I'd like to see this fixed. Things are completely flawless with Einstein@Home.

These problems are not on our end. It's mostly NOT related to the user's settings or the version of BOINC we are using. I realize the project is run by volunteers, but you need to accept there is a problem elsewhere.

Well, downloads are a bit slow and there are a lot of 'shorties', making it hard building
a cache.
They arrive, get crunched and send back.........
Nothing wrong with that.
[ADDED]
Us, (multi) GPU users, tend to be verrrrrryyyy hungrrrryyyyy and are
hard to satisfy....;-)

---

Well, downloads are a bit slow and there are a lot of 'shorties', making it hard building
a cache.
They arrive, get crunched and send back.........
Nothing wrong with that.

Slow is an understatement, though. I'm getting absolutely nothing like other people are posting here. There are probably many others who aren't getting anything and who don't post in the message boards. I didn't change anything at all, and one day, I stopped getting more work units to do. I eventually ran out of stored work units and had a hell of a time uploading the completed ones. One time some showed up in the transfers tab, but were impossible to download. After that happened I downgraded to 6.10.60 as suggested, but did not help.

Well, downloads are a bit slow and there are a lot of 'shorties', making it hard building
a cache.
They arrive, get crunched and send back.........
Nothing wrong with that.

Slow is an understatement, though. I'm getting absolutely nothing like other people are posting here. There are probably many others who aren't getting anything and who don't post in the message boards. I didn't change anything at all, and one day, I stopped getting more work units to do. I eventually ran out of stored work units and had a hell of a time uploading the completed ones. One time some showed up in the transfers tab, but were impossible to download. After that happened I downgraded to 6.10.60 as suggested, but did not help.

As long as keep returning valid results and enough work, when was that?, is
available and bandwidth, isn't maxxed out, it all works, but over a Million
(hungry) host's keep hammering at the 'front-door' making it hard to supply
every host equilly, in a particular time.

Sometimes I see a large DLoad qeue, no movement at all, it starts DLoading
cause waiting time is over followed by a DLoad, or 'clicking a (MB) WU, brings
it in motion, biggest problem persists, maxxed out bandwidth.IMO

About BOINC 6.10.58/59/60 and 6.12.33/34, probably in the change log
as well, could it be handling UP & DLoads, differently, f.i. shorter times between
requests. Little users would spend 24 hours, watching BOINC's every move :)

I like to try the latest versions, BOINC wise, more Anomous Accellerated Devices, i.e. the latest and very excellent v0.39 app.'s for ATI and NVidia, tip to the hat, for this beautiful app..
Took some time on ATI hosts, but certainly 'pays off'. Ehh, sliding off topic.......

[ADDED]
I had 1 MB WU, running for 24 hours, stuck @ 96.22%. Since there are little
MB WUs available, I've set ATI MB app. to 1 per device, pulse_find 1, to
'counter-act' a too long run time, or -177 error. But load is sufficient to keep
the core clock up @ 850MHz., VRAM 4800MHz, in a worst case it drops tru 300
to 157MHz. Memory speed decreases, with the same %-steps.

---

Duane wrote:

...
I realize the project is run by volunteers, but you need to accept there is a problem elsewhere.

See http://setiathome.berkeley.edu/forum_thread.php?id=36108&nowrap=true#505891 for Eric Korpela's explanation of funding and pay rates for the staff. That's from about 5 years ago, the situation has not improved since. The GPU User's Group has recently made it easier for some participants to donate, I hope many will consider whether keeping the project alive is a worthwhile use of some money.

                                                                   Joe

to the tune of "Here We Go Again"

OMG, I think I hurt myself falling off the chair, that is way too funny :)

---