Quite true, turnaround time is what counts and the BOINC database keeps that statistic for each app version on a host. The available BOINC feature uses that plus consecutive valid and having a quota at least equal to the basic setting of 100 to judge. The documentation is slightly out of date, but gives a reasonable overall view. Reading the code in sched_send.cpp and a few other source files gives accurate current info on how it works.

Having that feature turned on would of course be an additional load on the Scheduler processes, but with improved servers that may become feasible. What actual settings would be appropriate would take some thought, the general advice that about 25% of hosts ought to be considered reliable seems sensible. If set too tight, reissue tasks might occupy positions in the Feeder queue for too long.

Could we kill two birds with one stone and have two feeder queues in different instances of the scheduler? Ideally one with a big queue for the relable hosts that would include resends and a second with a smaller queue for the others.

LHC@Home 1.0 uses reliable hosts for resends (quick turnaround and high validated tasks) but I don't know how that's going for them, I've not read the forums for a while. But I do have a problem with two schedulers. Inevitably that would be seen as one for the high crunchers and one for the low crunchers. Besides, as has been said the servers are doing enough already, and that's quite a strain on them already despite the 100MB limit. The tasks will be done sooner or later, whether it's a day or months is irrelevant.
And to stay on topic, anybody have any idea on how many astropulses are flying around? I wouldn't mind a few :)

[edit]that's 5.05's I'm yakking about :)[/edit]

---

Member of the People Encouraging Niceness In Society club.

And to stay on topic, anybody have any idea on how many astropulses are flying around? I wouldn't mind a few :)

[edit]that's 5.05's I'm yakking about :)[/edit]

Yeah, just look at the server status page. The AP numbers have not been updated to include v6 yet. :D

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

And to stay on topic, anybody have any idea on how many astropulses are flying around? I wouldn't mind a few :)

[edit]that's 5.05's I'm yakking about :)[/edit]

Yeah, just look at the server status page. The AP numbers have not been updated to include v6 yet. :D

Aye, not updated and from the crickets it looks like astropulses V6 are being split like 5.05's. Hold on to yer socks :)

---

Member of the People Encouraging Niceness In Society club.

Making steady progress here.

Looking at cricket, anybody feels like jinxing it?

---

I'm not the Pope. I don't speak Ex Cathedra!

Looking at cricket, anybody feels like jinxing it?

I'll do it. QUACK! Still about 3 hours away from that, but it's too late now. :p

Also of note, I started my stats spreadsheet back up since we're over on v6 now. aside from stock not reporting how much was blanked, it will provide a good baseline to see how much better the optimization is compared to stock.

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

Looking at cricket, anybody feels like jinxing it?

I'll do it. QUACK! Still about 3 hours away from that, but it's too late now. :p

Also of note, I started my stats spreadsheet back up since we're over on v6 now. aside from stock not reporting how much was blanked, it will provide a good baseline to see how much better the optimization is compared to stock.

CPU? on benches between 2x-3x with another 20-25% extra on AVX.
r548/r555 is also faster than r409, but I can't find the values in a hurry.

---

I'm not the Pope. I don't speak Ex Cathedra!

Making steady progress here.

Looking at cricket, anybody feels like jinxing it?

No jinxes please....
All is going very well here in MB land.
Even with the new AP rollout, all boxen are up to their limits and keeping topped off.

Kitties chase away the duckens.

---

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

Kitties chase away the duckens.

Could the kitties please chase away the geesen that crappen all over all the sidewalken?

---

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

Kitties chase away the duckens.

Could the kitties please chase away the geesen that crappen all over all the sidewalken?

Dunno if it's the duckens or the geesen, but something crappen on the servers.
Although the cricket is still going full bore, and I can't see anything amiss on server status, suddenly none of my rigs have been issued a single MB task for about 45 minutes.

---

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

And then it came back, but you can see the chink in the Cricket graph.

---

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

I was gonna say, I just got 25 MB's in one hit... Maybe they had to reboot something?

---

All it took was one mention of that yellow thing.

---

10 MB-splitters and 12 AP-splitters are working now together. Are they testing new servers?

---

I'll let my computers run here and on Beta, and whatever they get they will crunch. I expect my RAC to crash, but as long as I have a high enough RAC to post on the boards, it's fine with me.

Your (if you really are Sten-Arne; this new attitude is cause for doubt) RAC is higher than mine, so you should be okay.

---

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

10 MB-splitters and 12 AP-splitters are working now together. Are they testing new servers?

Where do you get that information? Server status page shows AP1-6 (6), and MB 6,7,10,11,12 (5).

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

10 MB-splitters and 12 AP-splitters are working now together. Are they testing new servers?

Where do you get that information? Server status page shows AP1-6 (6), and MB 6,7,10,11,12 (5).

Obviously those new splitters are not counting at the status page as well as APv6 tasks, eh?
Look at the "channels in progress": 10 MB & 12 AP, and so at the graphical "Splitter status"

---

10 MB-splitters and 12 AP-splitters are working now together. Are they testing new servers?

Where do you get that information? Server status page shows AP1-6 (6), and MB 6,7,10,11,12 (5).

Obviously those new splitters are not counting at the status page as well as APv6 tasks, eh?
Look at the "channels in progress": 10 MB & 12 AP, and so at the graphical "Splitter status"

Of course if one is on the anonymous platform, the server now says one does not have a v6 AP app in the app_info.xml file.

SETI@home 3/15/2012 12:27:40 PM Message from server: Your app_info.xml file doesn't have a usable version of AstroPulse v6.

---

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

Wow!

Synergy finally got to see some action and now there's no stopping her!

Of course if one is on the anonymous platform, the server now says one does not have a v6 AP app in the app_info.xml file.

No, no, no! I mean that AP 6.01 tasks which was splitted and awaiting for crunchers or even sent to them, doesn't counting now at the server status page, there are only old AP 5.05 tasks.
And new splitters doesn't counting there too.

---

I think the listed splitters ARE the new v6 ones. No more 505's will ever be split. I have seen in the past where "channels in progress" are more than the number of splitters. It is actually somewhat common.

I don't believe it means there are more splitters than what are listed, it may just be that if one channel is taking too long or has a lot of I/O wait, that particular instance may move on to another channel, so two channels appear as active even though only one process is working on them.


of course, Occam's Razor would suggest that there are indeed a lot of splitters now but the list hasn't been updated to show them.

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)