A statistic to consider is that there are fewer than 4000 hosts with RAC above 2600. IOW, less than 1.5% of the ~279000 active hosts should be affected by a fixed 100 quota daily even if the supplied work were almost all VHAR shorties.

I just want to point out that you are making an assumption about the value of those 4000 hosts. A better stat would be: what the overall "crunch" from those 4000 are in comparison to those below 2600?

I made no assumption, merely gave some numbers for your consideration. I don't even know if project staff looked at that perspective as a factor in whatever settings they're using.

A few months ago I did study the contribution of top hosts, at that time the top 10000 hosts were doing about 39% of the work. That could have shifted since, but given a total RAC of about 73 million the average host RAC is around 260. There are roughly 60 thousand above average hosts doing half the work, 219 thousand below average also doing half the work.

Obviously a host with 26000 RAC is doing 10 times as much work as a host with 2600 RAC, and those of you who have high ranked hosts are very valuable. When the bugs have been worked out of the new server code, the intent is that any host producing consistently valid work will not be limited by a quota. Getting the bugs worked out is likely to be painful, and I suspect may never be totally completed.

                                                                 Joe

A statistic to consider is that there are fewer than 4000 hosts with RAC above 2600. IOW, less than 1.5% of the ~279000 active hosts should be affected by a fixed 100 quota daily even if the supplied work were almost all VHAR shorties.

I just want to point out that you are making an assumption about the value of those 4000 hosts. A better stat would be: what the overall "crunch" from those 4000 are in comparison to those below 2600?

I made no assumption, merely gave some numbers for your consideration. I don't even know if project staff looked at that perspective as a factor in whatever settings they're using.

A few months ago I did study the contribution of top hosts, at that time the top 10000 hosts were doing about 39% of the work. That could have shifted since, but given a total RAC of about 73 million the average host RAC is around 260. There are roughly 60 thousand above average hosts doing half the work, 219 thousand below average also doing half the work.

Obviously a host with 26000 RAC is doing 10 times as much work as a host with 2600 RAC, and those of you who have high ranked hosts are very valuable. When the bugs have been worked out of the new server code, the intent is that any host producing consistently valid work will not be limited by a quota. Getting the bugs worked out is likely to be painful, and I suspect may never be totally completed.

                                                                 Joe

That is very encouraging news indeed! I will help how ever I can, and if patience is called for, then I will not run out no matter what. :)

Steve

---

Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website

A statistic to consider is that there are fewer than 4000 hosts with RAC above 2600. IOW, less than 1.5% of the ~279000 active hosts should be affected by a fixed 100 quota daily even if the supplied work were almost all VHAR shorties.

I just want to point out that you are making an assumption about the value of those 4000 hosts. A better stat would be: what the overall "crunch" from those 4000 are in comparison to those below 2600?

I made no assumption, merely gave some numbers for your consideration. I don't even know if project staff looked at that perspective as a factor in whatever settings they're using.

A few months ago I did study the contribution of top hosts, at that time the top 10000 hosts were doing about 39% of the work. That could have shifted since, but given a total RAC of about 73 million the average host RAC is around 260. There are roughly 60 thousand above average hosts doing half the work, 219 thousand below average also doing half the work.

Obviously a host with 26000 RAC is doing 10 times as much work as a host with 2600 RAC, and those of you who have high ranked hosts are very valuable. When the bugs have been worked out of the new server code, the intent is that any host producing consistently valid work will not be limited by a quota. Getting the bugs worked out is likely to be painful, and I suspect may never be totally completed.

                                                                 Joe

That is very encouraging news indeed! I will help how ever I can, and if patience is called for, then I will not run out no matter what. :)

Steve

Yes indeed It is.

---

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

I find I now have the necessary ingredients to actually do some observation, research and thinking.

A host (3751792) where I can see the full result history, and which has reached to 100 quota. I also run extended logging, so I haven't lost any of the log file history.

I last received any work, and first received the '100 quota' message, at 12 Jun 2010 05:40 UTC. Counting back 100 tasks from that takes me to work issued at 11 Jun 2010 07:07 UTC. The task before that was issued 11 Jun 2010 06:48 UTC.

So far so good. My understanding is that quota is supposed to reset at midnight, local server time. So my task timings are consistent with a quota reset at 11 Jun 2010 07:00 UTC (midnight, Berkeley time), and 100 tasks (the current limit) being issued within 24 hours of that reset.

The previous day, I only received 57 tasks. This host has a 9800GT GPU card, and I only fetch SETI for GPU - other projects run on the CPUs - so around 60 is a typical daily throughput. It seems that this is sustainable.

But yesterday's 100 contained a significant number of VLARs, which get rescheduled. So the host needed more GPU work to replace what had been rescheduled, and that's what took it over 100.

The problem is, I'm still getting the '100 quota' message, most recently at 12 Jun 2010 09:15 UTC - that's over two hours after the quota should have reset (since then, the messages have been 'no work available').

So, the observed behaviour seems to be: if you have not reached daily quota, quota is reset daily at 07:00 UTC and you can carry on ad infinitum. But if you have reached quota, the daily reset mechanism breaks and you're stuck until they find the bug.

Observations? Corroborations?

Back on topic. This same host got

SETI@home 14/06/2010 19:23:38 Message from server: (reached daily quota of 100 tasks)

ten minutes ago. The last time it downloaded work remains 12 Jun 2010 5:40:01 UTC. That's one heck of a long 'daily'.

I have hundreds of results uploading, and haven't got a new task in days (I think June 10th was my last wu download). Once my reporting cycle is complete, I'll post my results (ie, did I reach my quota without getting anything or not etc).

---

Back on topic. This same host got

SETI@home 14/06/2010 19:23:38 Message from server: (reached daily quota of 100 tasks)

ten minutes ago. The last time it downloaded work remains 12 Jun 2010 5:40:01 UTC. That's one heck of a long 'daily'.

Yup, getting the same BUT only when using an app_info. With the app_info - even naming stock apps and info direct from the client_state.xml get all the errors - 'daily quota reached' and 'no usuable version of setiathome_enhanced' included.
Stopped Boinc renamed the app_info so it wasn't picked up and restarted Boinc - had 13 cuda(23) tasks download straight away then back to the no work available message?

Edit: Have just downloaded another 100 tasks (both cuda, cuda(23) and CPU) without and app_info.xml in place. Ran the Lunatics installed to test and now back to:
14/06/2010 20:06:29 SETI@home Message from server: No work sent
14/06/2010 20:06:29 SETI@home Message from server: No work is available for Astropulse v5
14/06/2010 20:06:29 SETI@home Message from server: Your app_info.xml file doesn't have a usable version of SETI@home Enhanced.
14/06/2010 20:06:29 SETI@home Message from server: (reached daily quota of 100 tasks)
Anyone else still having downlaod issues with an app_info?

---

Hi!

Once my PCs managed to upload, I could check again.
Several slower boxes with optimized clients and older boinc managet do download,
while two big CUDA boxes could not (Daily limit of 100, while nothing was DLed for 2 days). All PCs running linux.

I guess "100 per day" might indicate some other error?

Hopefully, it will get fixed soon. Or good tip pops up, how to hack around.

BR
Iztok

Well, finished my upload. Requested an update. Client requested GPU tasks. Got the message "Message from server: (reached daily quota of 100 tasks). Noything downloaded. I have work to continue 'till about the end of Thursday, so I can;t try anything until then. In the meantime, If I get a different (or additional) message, I will post here otherwise the assumption should be that my situation hasn't changed.

---

[quote]

Anyone else still having downlaod issues with an app_info?

Yup... Can't get any work for my Fermi too...
It worked perfect the first day I installed the card and then the strange message came...

My Fermi is idle since it ran out of WU's...
About 3 days now?

less than 1.5% of the ~279000 active hosts should be affected by a fixed 100 quota daily even if the supplied work were almost all VHAR shorties.

The statement you offer seemed like a judgement on the value of the contribution from those 1.5% from the way it was presented.

I withdraw my observation.

BTW, I just uploaded for the first time since Saturday afternoon and received a fullsome set of 100 CUDA WUs. At the pace I am crunching them I will be back to running on empty in a few hours time - a fraction of my capacity to crunch.

Still don't quite understand the value of where this is going but if you say that patience will reward us, well idle GPUs are patient if nothing else (well, quiet too).

Regards,

---

less than 1.5% of the ~279000 active hosts should be affected by a fixed 100 quota daily even if the supplied work were almost all VHAR shorties.

The statement you offer seemed like a judgement on the value of the contribution from those 1.5% from the way it was presented.

I withdraw my observation.

BTW, I just uploaded for the first time since Saturday afternoon and received a fullsome set of 100 CUDA WUs. At the pace I am crunching them I will be back to running on empty in a few hours time - a fraction of my capacity to crunch.

Still don't quite understand the value of where this is going but if you say that patience will reward us, well idle GPUs are patient if nothing else (well, quiet too).

Regards,

My rig uploaded as well, but won't download a single WU. I am happy you got some work. :)

Steve

---

Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website

That is so strange Steve. Something you might try is to completely stop BOINC and then restart it. I took BOINC offline just to stop it continually knocking on SETI's server. Putting it back online it immediately uploaded and downloaded. Might be nothing but might help you give everything a nudge.

Regards,

---

Hopefully, it will get fixed soon. Or good tip pops up, how to hack around. BR Iztok

Quota stays for me too, but we could ask our friend RottenMut. He managed to get 226 fresh clean MB tasks for his Top Rig (yes I count each and every one of them :)
Detaching and reataching seems to work, but only when your rigs are dry clean.
It doesn't work for me, not while I have 7 days worth of AP units. Hard to get, terrible waste if trashed :)

---

Who the hell is General Failure and why is he reading my harddisk?¿

He managed to get 226 fresh clean MB tasks for his Top Rig (yes I count each and every one of them :)

You just made me go back and count.

I actually got 113 CUDA tasks, not the 100 quota-limited download. I am aldo getting the "no jobs available" message and no longer the quota one.

Has something changed?

Regards,

---

The quota should be per app version, 100 for CPU Astropulse_v505, 100 for CPU Seti_enhanced, 100 for Cuda_Fermi, 100 for Cuda32, 100 for Cuda, 100 for ATI Astropulse_v505, etc.

If you're getting the message 'Message from server: Your app_info.xml file doesn't have a usable version of SETI@home Enhanced.',
perhaps the server has sent you enough work for Cuda, but now wants to send you work for Cuda32, but that app_class isn't in your app_info, so can't, hence the message,

see my Beta host's Application info page for a better understanding,

or Richard's Beta hosts, for hosts that have had their quota's upped because they've had tasks validated,

Claggy

He managed to get 226 fresh clean MB tasks for his Top Rig (yes I count each and every one of them :)

You just made me go back and count.

I actually got 113 CUDA tasks, not the 100 quota-limited download. I am aldo getting the "no jobs available" message and no longer the quota one.

Has something changed?

Regards,

Not necessarily. Both of these behaviours have been noted before.

1) Overshoot. If you've downloaded, say, 98 tasks (below quota), and you're still asking for substantial quantities of work, you might still get allocated 15 tasks on the next request. That would make your 113. It shouldn't happen - the quota system should limit you to no more than two on that final request - but it does. Been reported as a bug.

2) No jobs available. This message seems to trump all others. If you get it, none of the other messages appear. But next time there is work available, one or more of the other messages are likely to appear, giving you a plethora of reasons why you might not be getting it.

Ah!

This is the only interesting message I'm getting (beyond the ones noting the traffic as I complete and upload):

14/06/2010 7:05:56 PM SETI@home Message from server: Project has no jobs available

As per Richard that might explain things.

All those downloaded wre CUDA so I was thinking I'd managed to evade the quota, or, it has again been relaxed.

Sounds like neither of these.

Regards,

---

The quota should be per app version, 100 for CPU Astropulse_v505, 100 for CPU Seti_enhanced, 100 for Cuda_Fermi, 100 for Cuda32, 100 for Cuda, 100 for ATI Astropulse_v505, etc.

Please advise where you received this info from. Did it come from SETI??

Are the SETI servers then going to keep track of the VLAR work and assign those wu only to the cpu's? So then we can stop using the rescheduler!

---

Boinc....Boinc....Boinc....Boinc....

Well, I was task empty, and I tried detaching and reattaching, and got several Fermi WU's. I will poke around a bit to see if I get some CPU tasks. I would like to fill my CPU as well, so my chiller won't cycle. It feels great to do some work. The shorties I am getting are finishing in two minutes.

Steve

---

Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website

The cork continues to come out of the bottle.

I tried abusing some buttons (all in the name of investigative science you understand) and caused the following

14/06/2010 7:12:24 PM SETI@home Sending scheduler request: To fetch work.
14/06/2010 7:12:24 PM SETI@home Requesting new tasks for GPU
14/06/2010 7:12:26 PM SETI@home Scheduler request completed: got 0 new tasks
14/06/2010 7:12:26 PM SETI@home Message from server: Project has no jobs available
14/06/2010 7:13:34 PM SETI@home update requested by user
14/06/2010 7:13:38 PM SETI@home Sending scheduler request: Requested by user.
14/06/2010 7:13:38 PM SETI@home Requesting new tasks for CPU and GPU
14/06/2010 7:13:40 PM SETI@home Scheduler request completed: got 17 new tasks

The fresh 17 are non-CUDA so I think Richard's observation above fits.

[edit]

aaaaaanddd.... the dreaded quota message now returns.....sigh

14/06/2010 7:20:02 PM SETI@home Scheduler request completed: got 0 new tasks
14/06/2010 7:20:02 PM SETI@home Message from server: No work sent
14/06/2010 7:20:02 PM SETI@home Message from server: No work is available for Astropulse v5
14/06/2010 7:20:02 PM SETI@home Message from server: (reached daily quota of 100 tasks)
14/06/2010 7:20:17 PM SETI@home Sending scheduler request: To fetch work.
14/06/2010 7:20:17 PM SETI@home Requesting new tasks for CPU and GPU
14/06/2010 7:20:20 PM SETI@home Scheduler request completed: got 0 new tasks
14/06/2010 7:20:20 PM SETI@home Message from server: Project has no jobs available

[/edit]

Regards,

---