Is anyone else having problem getting new work?

Nick

---

Yea, I have a wack of 59:30 wu's, only about 12 hours worth. Hopefully, more work will get split soon.

Is anyone else having problem getting new work?

Nick

Looks like WUs are being turned it at ~15/sec, but the splitters are only running ~13/sec. At that rate there'll never be enough to satisfy the demand.

---

"Gentlemen, there are only two types of naval vessels..........Submarines, and Targets" -- U.S. Navy Submarine SONAR Instructor.

Same here. Cant get new work.

---

Very hit and miss. I just had one of my rigs get through to the sceduler and it snagged a handful of WUs about 15 minutes ago. However it still has not been able to complete any of the downloads.
The way the Cricket Graph shows the bandwidth being hammered, it's a wonder anything including the forums is moving at all. I don't recall seeing such bandwidth being sustained at that level for a while. Glad Matt is keeping an eye on things while he is on holiday, cuz it wouldn't surprise me if something goes poof.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

I don't recall seeing such bandwidth being sustained at that level for a while. Glad Matt is keeping an eye on things while he is on holiday, cuz it wouldn't surprise me if something goes poof.

Maybe we all need to stay off the forum for 12 hours or so to decrease the bandwidth on the server. I for one will, and have stopped my reports for a 24 hour period on both rigs.

Just a thought...

I don't recall seeing such bandwidth being sustained at that level for a while. Glad Matt is keeping an eye on things while he is on holiday, cuz it wouldn't surprise me if something goes poof.

Maybe we all need to stay off the forum for 12 hours or so to decrease the bandwidth on the server. I for one will, and have stopped my reports for a 24 hour period on both rigs.

Just a thought...

I just suspended SETI for now, it's got 5 workunits wanting to upload (for the last 2 hours) 22 waiting to run (and that will take my system about 10 hours)

I'll just leave it doing TSP for now since their GA_ workunits are looking like 10 hour ones anyway.

I'll start SETI back up when things settle down, the workunits have a deadline of 31st DEC, so plenty of time to let things relax a bit :)

---

LOL....
My phased Penny actually got 3 of the current WUs assigned.
One took 4.98 seconds to complete.
The second took 630.8 seconds.
And the third has been trying to download for the last 15 minutes. It'll probably take less time than that to crunch it when the download finally gets through.

EDIT...
I was right...after the download, the third took 675.05 seconds to crunch. Less time that the download took.
It's easy to see why the splitters are not gaining any ground.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

heh just looking at the server status page shows why.

56,127 Results recived in the last hour, thats about 15.59 per second
And they are only creating 18.68 results a second.

No wonder they cant keep up!

---

Is anyone else having problem getting new work?

Nick

No new work for the last 20 hours. All my caches are exhausted, so my machines are idle right now. One is backed off by 2+ hours waiting to d/l the last 40 KiloBytes of 1 unit. All requests for work gets the reply: got 0 new tasks.
Server status shows all green but one, but no work units are ready.

Last report from staff is from 1800 UTC on Dec 21. Seems they have a nasty one here :(
I am sure they are doing their best, but hey, it's Christmas.

We'll idle for a few days and You go and spend the Christmas with Your loved ones. :)

Merry Christmas to You all.

ChrisD

---

Just ran out of SETI, this is the most recent msg on the client:

12/23/2007 11:45:54 PM|SETI@home|Sending scheduler request: Requested by user. Requesting 0 seconds of work, reporting 9 completed tasks

Why its requesting 0 seconds is prolly due to the scheduler (not sure)

But, I keep a 1 day cache anyhow and SETI gets 25% - one core - on the new rig (quaddy + large cache + 56k connection = bad)

I have enough Spinhenge & TSP to cover the downtime

Just ran out of SETI, this is the most recent msg on the client:

12/23/2007 11:45:54 PM|SETI@home|Sending scheduler request: Requested by user. Requesting 0 seconds of work, reporting 9 completed tasks

Why its requesting 0 seconds is prolly due to the scheduler (not sure)

But, I keep a 1 day cache anyhow and SETI gets 25% - one core - on the new rig (quaddy + large cache + 56k connection = bad)

I have enough Spinhenge & TSP to cover the downtime

If Boinc is not requesting more work from Seti, it is probably because it needs to run more work from your other projects to maintain the resource share that you set. That is a normal occurrence when running more than one project. Boinc keeps track long term of how much time each project has had access to the CPUs, when the share has been rebalanced it will request more work from Seti.

---

Boinc V7.2.42
Win7 i5 3.33G 4GB, GTX470

Started getting new wu's. Merry Christmas everyone.

---

If Boinc is not requesting more work from Seti, it is probably because it needs to run more work from your other projects to maintain the resource share that you set. That is a normal occurrence when running more than one project. Boinc keeps track long term of how much time each project has had access to the CPUs, when the share has been rebalanced it will request more work from Seti.

Resource share is 25% SETI, 25% TSP, 50% Spinhenge (1 core, 1 core, 2 cores). Both SETI/Spinhenge have been running since this rig went live (was 25/75), but dropped SH down for TSP. The balance works out in the end and aside from server glitches, haven't had an issue running out on a project... Frankly, I'm not too concerned since its not really for the credits (though they are nice) but the end result of the project. If I ever get highspeed in my area, I'll change the share, but for now, that's all my connection can handle as far as SETI goes.

But like the OP, it fixed itself this morning... one -9 overflow, 12 in cache :)

12/23/2007 8:53:44 PM|SETI@home|Sending scheduler request: To fetch work. Requesting 95223 seconds of work, reporting 12 completed tasks

^^3 were pending upload

Do I recall correctly that some of the Boinc clients will not request new work if there are downloads pending? I am wondering if some hosts were not asking for new work because for the last few days, downloading work was painfully slow, and failed downloads would sometimes time out for hours before retrying. Perhaps stuck downloads were part of the reason for no new work?

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Do I recall correctly that some of the Boinc clients will not request new work if there are downloads pending? I am wondering if some hosts were not asking for new work because for the last few days, downloading work was painfully slow, and failed downloads would sometimes time out for hours before retrying. Perhaps stuck downloads were part of the reason for no new work?

I had three hosts that ran out of work, 1 pending download on each, they all got a ton of work after I aborted the pending DL.

-Dave

---

Do I recall correctly that some of the Boinc clients will not request new work if there are downloads pending? I am wondering if some hosts were not asking for new work because for the last few days, downloading work was painfully slow, and failed downloads would sometimes time out for hours before retrying. Perhaps stuck downloads were part of the reason for no new work?

I had three hosts that ran out of work, 1 pending download on each, they all got a ton of work after I aborted the pending DL.

-Dave

Ahh...so it appears that my recollection may have been correct.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Ahh...so it appears that my recollection may have been correct.

Yes, and what's interesting is that I tried getting more work manually with them all before aborting and got the
Not requesting new work or reporting tasks
message... Had me baffled for a little while until I noticed the DL pending and checked how long it had been "in process". I think they were all at something like 24+ hours in the "elapsed time" column.

-Dave

---

I'm using BOINC v5.10.13. It won't request new work while any downloads are in the "Retry in hh:mm:ss" state, but it will request work while WUs are "Downloading" or "Download pending".

Solution - open the transfers tab in grid view, mouse over the list to select them all, and hit 'retry'. You may have to repeat after 21 seconds (remember them?), because I think BOINC only thinks about requesting new work once a minute, but you should see a scheduler work request soon enough. Doesn't mean you have to abuse the buttons to complete the download, doesn't guarantee you'll actually get any new work, but it does mean that you don't have to abort anything.

I'm using BOINC v5.10.13. It won't request new work while any downloads are in the "Retry in hh:mm:ss" state, but it will request work while WUs are "Downloading" or "Download pending".

Solution - open the transfers tab in grid view, mouse over the list to select them all, and hit 'retry'. You may have to repeat after 21 seconds (remember them?), because I think BOINC only thinks about requesting new work once a minute, but you should see a scheduler work request soon enough. Doesn't mean you have to abuse the buttons to complete the download, doesn't guarantee you'll actually get any new work, but it does mean that you don't have to abort anything.

Simpler solution update to a newer version of Boinc!!!

---

Boinc V7.2.42
Win7 i5 3.33G 4GB, GTX470