I'm seeing a larger than normal number of rework tasks (..._2, 3, or 4) in 3 different frequency ranges; so they might be out of unstarted blocks of data. With locality scheduling trying to minimize the number of people working on each block of data it's possible that if you've got CPU-BRP4 tasks allowed the scheduler is giving you those instead of something that would require a large amount of fresh data to be downloaded.
I deselected "Binary Radio Pulsar Search (Arecibo)" - and now I often don't get any new work units (which is good for "World Communitx Grid").
(FWIW I deselected "If no work for selected applications is available, accept work from other applications?").
/me wonders why there's often an empty queue fot these 2 projects:
I'm still getting s6 CPU tasks as normal. The S6 and BRP4-CPU tasks have different DCF values; it's possible that when your client ended up with a bigger than normal share of the latter it resulted in your project debt values being shifted due to the BRP4's taking longer than normal and E@H's share of time getting too high. If so it should balance itself out over the next few days as WCG catches up.
The S6 and BRP4-CPU tasks have different DCF values; it's possible that when your client ended up with a bigger than normal share of the latter it resulted in your project debt values being shifted due to the BRP4's taking longer than normal and E@H's share of time getting too high. If so it should balance itself out over the next few days as WCG catches up.
No more S6 LineVeto WUs?
)
I'm seeing a larger than normal number of rework tasks (..._2, 3, or 4) in 3 different frequency ranges; so they might be out of unstarted blocks of data. With locality scheduling trying to minimize the number of people working on each block of data it's possible that if you've got CPU-BRP4 tasks allowed the scheduler is giving you those instead of something that would require a large amount of fresh data to be downloaded.
I deselected "Binary Radio
)
I deselected "Binary Radio Pulsar Search (Arecibo)" - and now I often don't get any new work units (which is good for "World Communitx Grid").
(FWIW I deselected "If no work for selected applications is available, accept work from other applications?").
/me wonders why there's often an empty queue fot these 2 projects:
Gravitational Wave S6 LineVeto search
Gamma-ray pulsar search #1
My BOINC client has version 7.0.39 and runs w/o a GPU, just a 4-core CPU (i5).
FGRP has been on hold for a
)
FGRP has been on hold for a while. It will resume after the initial results are published and a new app version is written.
http://einsteinathome.org/goto/comment/110906
I'm still getting s6 CPU tasks as normal. The S6 and BRP4-CPU tasks have different DCF values; it's possible that when your client ended up with a bigger than normal share of the latter it resulted in your project debt values being shifted due to the BRP4's taking longer than normal and E@H's share of time getting too high. If so it should balance itself out over the next few days as WCG catches up.
RE: The S6 and BRP4-CPU
)
Well, here are my stats:
IS this normal ?
RE: IS this normal ? If
)
If you've deselected CPU-BRP4, then yes. There aren't many S6LV-Tasks around, as stated above.