It depends on which version of the Cricket graph you link to....
Graphs for gigabitethernet2_3 is inside out....but
Graphs for gigabitethernet5_1 shows things as they are from the outside world....in is in and out is out.

Never knew about the other one either...thanks, Mark!!

---

Clk2HlpSetiCty:::PayIt4ward

Is anybody more demoralized about this AP transition than I am?

Not only have the servers been damned for 3-4 days (so far), and a large number of people are reporting no work, and the network connection is in overdrive meaning we can't return results reliably, and ...; there is also a large number of AP wu's listed as being sent out but not appearing on the client machines (I think I have at least two like that). Plus, I don't want to run AP until they can figure things out-- and there appears to be a lot of things to figure out. But even that safeguard was never tested.

Ever feel like we all are playing Rodney Dangerfields' character? In a sense I feel like that piece of gum on the sole of the shoe on a hot summer day that just won't scrape off no matter how hard you try or how badly it's treated.

My faster machine appears to have some work, but my other one is practically out, and has downloaded several five-hour Einsteins. The Message column of the latter machine's Boinc Manager adds two lines per second. Kinda funny!

---

It depends on which version of the Cricket graph you link to....
Graphs for gigabitethernet2_3 is inside out....but
Graphs for gigabitethernet5_1 shows things as they are from the outside world....in is in and out is out.

Never knew about the other one either...thanks, Mark!!

I forget who turned me on to the 5_1 graph, but it gives a much truer picture of what is going in and out of the servers to us....

Can you imagine how much worse things would be if their bandwidth ceiling was still around 60Mbit/second like it used to be?

---

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

Is anybody more demoralized about this AP transition than I am?

Not only have the servers been damned for 3-4 days (so far), and a large number of people are reporting no work, and the network connection is in overdrive meaning we can't return results reliably, and ...; there is also a large number of AP wu's listed as being sent out but not appearing on the client machines (I think I have at least two like that). Plus, I don't want to run AP until they can figure things out-- and there appears to be a lot of things to figure out. But even that safeguard was never tested.

Ever feel like we all are playing Rodney Dangerfields' character? In a sense I feel like that piece of gum on the sole of the shoe on a hot summer day that just won't scrape off no matter how hard you try or how badly it's treated.

The Problem? does it exsist or is it home made. I am at present running 8 AP 4.35 on two host over here on beta and they are stable. All AP4.35 that have been returned this month have all been validated state valid and or are awaiting wingpersons.

I think the credits could be a bit larger.

I get all the work I need over here and will wait for the AP 4.35 to arrive through the post up to now I have received no AP4.35 and have no intention of trying to get any. Stock apps and non O/C has always worked for me and I will continue to use them.
Michael

Is anybody more demoralized about this AP transition than I am?

We've seen this kind of problem before, and yet, we've never seen Astropulse on the main project until now.

The problem in the past have been "shorties" and in fact several posts on this thread talk about things like the turn-around time dropping, and going through workunits quickly.

... and the one thing we know about AP work units is that they take a LONG time. If there was a lot of astropulse out there, the turn around time would go up, not down.

If anything, it seems that we'd be better off right now if the AP rollout had been more aggressive -- or even if they turned off the Multibeam splitters completely and just split Astropulse for a little bit.

Oh, wait, can't do that because the folks running optimized apps would all run out of work.

Ultimately, I think we demand far more from the project than they ever promised to deliver -- and far more than we should expect given the staffing and funding.

---

Is anybody more demoralized about this AP transition than I am?

We've seen this kind of problem before, and yet, we've never seen Astropulse on the main project until now.

The problem in the past have been "shorties" and in fact several posts on this thread talk about things like the turn-around time dropping, and going through workunits quickly.

... and the one thing we know about AP work units is that they take a LONG time. If there was a lot of astropulse out there, the turn around time would go up, not down.

If anything, it seems that we'd be better off right now if the AP rollout had been more aggressive -- or even if they turned off the Multibeam splitters completely and just split Astropulse for a little bit.

Oh, wait, can't do that because the folks running optimized apps would all run out of work.

Ultimately, I think we demand far more from the project than they ever promised to deliver -- and far more than we should expect given the staffing and funding.

Yes a possible solution after outage today let in a few Clean AP4.35 WU on there own a Hundred or so let them be taken up and then restart MB again.

This may solve the problem of so many ghosts. and may solve a few of the other problems and also give a pointer as to where to start looking.

Michael

no work for me too .... what's going on ...

---

There's a bug on all those Cricket graphs...
The raw image sizes are actually 597 x 263 pixels, but the html scales them
to 595 x 255.. so the resulting images are not as crisp as they should be.

Well almost run dry on one machine (already dry on the other). Funny thing is the machine is supposed to hold a five day cache and is busy trying to upload finished work (having great difficulty) but is refusing to request new work. I guess that is due to upload problems.

I'm going to have to find it another source of work for the night when I go home in about 6 1/2 hours...

---

Never surrender and never give up. In the darkest hour there is always hope.

From what is said here, I must have got lucky yesterday at about 15:10 UTC and cleared uploads, as my quad downloaded 400 MB units before 16:20. That I think is full days quota.

Trouble is about 95% are all shorties, approx 14 mins. So that won't last long.

edit] Had hit 'retry comms' just before typing this post, cleared most of upload queue and downloaded 30+ MB units, but this time there are normal length, 40 - 50 min units.

Most of you should be getting some work soon......uploads are starting to go through after a number of tries, and once they clear, you should start to request and receive new work....YAY!!!

---

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

hi there,
can someone tells me , what this means?


WU true angle range is : 1.568089

how many spikes or gausians
must i find,
just to be an alien signal?

thanx

hi there,
can someone tells me , what this means?


WU true angle range is : 1.568089

how many spikes or gausians
must i find,
just to be an alien signal?

thanx

The angle_range [AR) comes from the movement of the telescope when the unit was recorded. Very Low AR is when the telescope is tracked so it is constantly looking at one point of the sky, Mid range AR=0.42 is when the telescope is stationary and movement of the earth is only tracking. Very High AR is when telescope is moving.
Gaussians processing is only done in mid range from AR = 0.23 to ~ 1.1.

You only need to find one signal, but that would have to be tested, and it would have to be detected more than once, over a period of time. Hence NTPCKR, see sticky.

Everything seems a-okay now. Uploads and downloads are all going through licketty-spit for me.

This last week has had so many ups and downs.........worst thing was that I just got my second machine up and going last week and was so keen to see how it would crunch SETI...then.....no work!!!

A BTW question........is it normal when you increase in website preferences, "Maintain enough work for an additional ***" to a value more than the default to get many many 'shorties'? I increased that value to '1' a couple of days ago and got, like, about 30 shortie WU's...............Before I increased that value I always had maybe 2-4 WU's in cache....each taking about 50mins of crunching..........Now I have about 20 shorties in cache...each taking about 16 minutes. Normal?

Everything seems a-okay now. Uploads and downloads are all going through licketty-spit for me.

This last week has had so many ups and downs.........worst thing was that I just got my second machine up and going last week and was so keen to see how it would crunch SETI...then.....no work!!!

A BTW question........is it normal when you increase in website preferences, "Maintain enough work for an additional ***" to a value more than the default to get many many 'shorties'? I increased that value to '1' a couple of days ago and got, like, about 30 shortie WU's...............Before I increased that value I always had maybe 2-4 WU's in cache....each taking about 50mins of crunching..........Now I have about 20 shorties in cache...each taking about 16 minutes. Normal?

The mix of WU's you get depends solely on the type that the splitters are splitting and putting into the RTS queue. Nothing at your end affects the mix in any way.

F.

---

^ thanks.