Results out in the field @ 01:30 UTC 2,260,425

Results out in the field @ 09:30 UTC 2,249,424

Results out in the field @ 10:00 UTC 2,247,061

Results returned and awaiting validation @ 01:30 UTC 4,003,383

Results returned and awaiting validation @ 09:30 UTC 2,478,632

Results returned and awaiting validation @ 10:00 UTC 2,476,702

So the Validation backlog is now just about caught up. I think it's usually 24 hours that Valid results are visible on user and host pages.

Out in the field is more interesting anyway to me. As long as this is above 0, there's still a chance to grab a new WU :)

The ones "In the field" have already been sent out, and being processed by users
Ready to send shows the likelihood of getting a new task.

The ones "In the field" have already been sent out, and being processed by users

But they might generate a new task if they time out, fail to validate, or are part of a workunit requiring extra checking.

Out in field / (results in last hour * 24) = about 20 days.
Stragglers after that for some time.

Just got three on my 8 core system. They are all GPU units. Nothing on my Workstation

---

Historian for the Defunct Riviera Hotel and Casino, Former Classic Seti@home user for Team Art Bell. Greetings from the High Desert!

Current result creation rate **	0/sec	1.8678/sec	32.4094/sec

A final hurrah?

Forget about... was a false alarm, maybe a ghost in the machine.

---

Strangely SETI is showing 90 tasks in progress for me, when I finished WUs 2-3 April!
What's that about?
Oh wait, I remember trying to update the optimised app on my 2nd rig & the WUs vanished! I went back to the std install but they never came back :(. Could they still be in there somewhere?

---

Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, DHEP@H.

Main rig - Ryzen 5 3600, 32GB DDR4 3200, RX 580 8GB, 500GB Samsung 970 Evo+, Win 10
2nd rig - i7 4930k @4.1 GHz, 16GB DDR2 1866, HD 7870 XT 3GB (DS), Win7 64bit

Strangely SETI is showing 90 tasks in progress for me, when I finished WUs 2-3 April!
What's that about?
Oh wait, I remember trying to update the optimised app on my 2nd rig & the WUs vanished! I went back to the std install but they never came back :(. Could they still be in there somewhere?

Probably you have a graveyard full of ghosts. Try to follow this procedure to recover them.
https://setiathome.berkeley.edu/forum_thread.php?id=84176

---

The op from that thread seems to have disappeared :(, any idea what the procedure was?

---

Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, DHEP@H.

Main rig - Ryzen 5 3600, 32GB DDR4 3200, RX 580 8GB, 500GB Samsung 970 Evo+, Win 10
2nd rig - i7 4930k @4.1 GHz, 16GB DDR2 1866, HD 7870 XT 3GB (DS), Win7 64bit

His procedure was complicated, and relied on very exact timing. Other procedures exist, which don't rely on timing, but are even more complicated.

The good news is that the Kittyman has reported a conversation with Eric, who said that the internal server-based resend mechanism was being turned on periodically, when the database load permitted it (and by extension, turned off again when the server got busy). So, just leave your computer running and connected to SETI, and in a state where it will periodically request new work - you might get lucky. If you happen to be passing by, hit 'update' for SETI every few hours, to keep it interested.

Edit - it's only the original opening post which has been deleted. Keith's procedure is re-stated in the middle, at post 2010025.

I think it might be time to reduce the client backoff to 3 or 5 minutes instead of 30, now that the servers have recovered from panic mode.
Hitting the Update button still works between scheduled requests
Obviously if this starts causing problems, then 30 or 15 might still be appropriate

The op from that thread seems to have disappeared :(, any idea what the procedure was?

do you have at least 20 or 30 tasks to process now?

Stephen

? ?

I think it might be time to reduce the client backoff to 3 or 5 minutes instead of 30, now that the servers have recovered from panic mode.
Hitting the Update button still works between scheduled requests
Obviously if this starts causing problems, then 30 or 15 might still be appropriate

. . If there is work and you request it less than the 30 mins after a previous request you will get that message and not get any work. So unless they reduce the server settings it is best to wait the 30 mins.

Stephen

:(

No, no tasks left, ran out a couple of days after they turned the taps off.
TBH I thought that was it & have moved onto for Rosetta for now. The only reason I brought this up is because I discovered it was still showing me on the website that I have 90 tasks.
That's when I realised when I tried to update the optimised app & the WUs vanished, that's it's probably those WUs, & maybe I still have them?
Anyway, I wasn't looking to get new WUs, but see if I can uncover the ones I've got.

Richard
I'll check out the link, thanks :)

---

Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, DHEP@H.

Main rig - Ryzen 5 3600, 32GB DDR4 3200, RX 580 8GB, 500GB Samsung 970 Evo+, Win 10
2nd rig - i7 4930k @4.1 GHz, 16GB DDR2 1866, HD 7870 XT 3GB (DS), Win7 64bit

Current result creation rate **	0/sec	1.8678/sec	32.4094/sec

A final hurrah?

Forget about... was a false alarm, maybe a ghost in the machine.

There have been 2 rather odd jumps in the Results out in the field numbers that show up in the graphs.

---

Grant
Darwin NT

No, no tasks left, ran out a couple of days after they turned the taps off.
TBH I thought that was it & have moved onto for Rosetta for now. The only reason I brought this up is because I discovered it was still showing me on the website that I have 90 tasks.
That's when I realised when I tried to update the optimised app & the WUs vanished, that's it's probably those WUs, & maybe I still have them?
Anyway, I wasn't looking to get new WUs, but see if I can uncover the ones I've got.

Richard
I'll check out the link, thanks :)

. . The only process I know that works is having several results to upload and trigger the ghost resend. It used to be limited to 20 ghosted WUs but that has been increased greatly and is at least 40 WUs at a time now, it may even be greater than that. But when you have nothing actually in progress to return there is no way that I know of. They will be resent when their deadlines expire, just not to the original host/volunteer.

Stephen

. . :(

I think it might be time to reduce the client backoff to 3 or 5 minutes instead of 30, now that the servers have recovered from panic mode.
Hitting the Update button still works between scheduled requests
Obviously if this starts causing problems, then 30 or 15 might still be appropriate

Nope.
The problem with the shorter time between requests is the BOINC manager goes in to extreme backoff mode that much sooner. With 20-30min between requests i only have to click Update every couple of hours instead of every 20min or so to keep it requesting work.

---

Grant
Darwin NT

Current result creation rate **	0/sec	1.8678/sec	32.4094/sec

A final hurrah?

Forget about... was a false alarm, maybe a ghost in the machine.

There have been 2 rather odd jumps in the Results out in the field numbers that show up in the graphs.

. . That was not a glitch, those resends were real, I got 15 of them on my Windows (slowest) box, nada on the other 2. The same old Murphy's Law rule is in force, the slowest machines get the work first :(

Stephen

:(

No, no tasks left, ran out a couple of days after they turned the taps off.
TBH I thought that was it & have moved onto for Rosetta for now. The only reason I brought this up is because I discovered it was still showing me on the website that I have 90 tasks.
That's when I realised when I tried to update the optimised app, & the WUs vanished, that it's probably those WUs, & maybe I still have them?
Anyway, I wasn't looking to get new WUs, but see if I can uncover the ones I've got.

Richard
I'll check out the link, thanks :)

. . The only process I know that works is having several results to upload and trigger the ghost resend. It used to be limited to 20 ghosted WUs but that has been increased greatly and is at least 40 WUs at a time now, it may even be greater than that. But when you have nothing actually in progress to return there is no way that I know of. They will be resent when their deadlines expire, just not to the original host/volunteer.

Stephen

. . :(

Damn, OK thanks for LMK Stephen.

---

Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, DHEP@H.

Main rig - Ryzen 5 3600, 32GB DDR4 3200, RX 580 8GB, 500GB Samsung 970 Evo+, Win 10
2nd rig - i7 4930k @4.1 GHz, 16GB DDR2 1866, HD 7870 XT 3GB (DS), Win7 64bit

Thanks for the updates Eric.

I still remember with great fondness our few days in San Fran and the guided tour of the lab and the Holy Grail of the original server closet. To this day DA never knew I sat in his chair (Ooooops!)

It was my greatest pleasure to buy you the Centurion server in memory of my dad that dies aged 101, not fully realising at the time how much it was so desperately needed.

Unlike some, I will not say goodbye, I will say Au Revoir. The set-up we have here is too good to go to waste. Get together with DA and get Nebula up and running, then we can have some second sift work to process.

My feelings exactly, thank you for so generously hosting me on my trip out there, and allowing me to tag along with you to the lab as well as the tours seeing it all, it was something I won't forget.

And my server gift in honor of my parents was defiantly a highlight for me as well, I hope that all of ours will serve you well in the upcoming processes that you are now beginning. Here's hoping that one day in the not too distant future, SETI 2.0 will arise from the ashes, even better than before, and we continue on with our quest to find ET.

---