> Einstein fix confirmed. Just update your general prefs and the next time
> it'll work. Click General pref, then Update. Don't bother changing
> anything.

Right.

Only now it seems you'd have to do that every time in order to get a response from the scheduler. But it still works

---

> > ....
> Siran and Misfit... i guess they made some changes to the scheduler that
> requires you to give it a little nudge. Oddly enough, if you update your
> general preferences for Einstein, then update the project, it all starts
> working again. :)
>
> Thanks to Rockenstein for the tip!
>
> Dig
>

Thanks Digger. Things are better again with E@H. Kind of a weird fix. Oh, thanks to Rockenstein too, whomever he may be. Thanks again....

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

> All outages get this sort of response for sure -- for me, planned and
> announced outages cause no bother at all -- any inconvenience is modest --
> especially if they are announced with some real lead time.
>
> Unplanned outages tend to irritate me a bit more -- they carry a bit more
> inconvenience.
>
> Unplanned extended outages (more than a day) bother me (more than some who
> don't care I suppose).
>
> And unplanned extended outages where the status reporting mismatches what is
> going on are troublesome to me.

Yet, these kinds of outages are planned as part of the overall project design. BOINC caches work because outages happen. It holds work until it can be uploaded because outages happen. Work continues during the outage.

So, my question is, why do people use words like "irritate" or "bother" or "troublesome" when according to your message the only real effect is a bump in RAC?

Because the technology works.

I don't care that much about outages because, at the very worst, it just changes what I'm crunching -- and if I cared about that, I'd increase my cache.

---

"BOINC caches work because outages happen. It holds work until it can be uploaded because outages happen. Work continues during the outage."

In SETI I could get a cache of units for my computer to work on when the site was down, but how does one do it in BOINC?

Thanks,

Nick

---

> In SETI I could get a cache of units for my computer to work on when the site
> was down, but how does one do it in BOINC?

Your Account link, View or Edit general preferences, Edit preferences.
Under Network USage, set connect every x days to the desired number. In my case setting it to 6 days gives me 4 days worth of work (BOINC estimates 4.5 hrs per Work Unit, i'm doing them in around 3hrs 15 min).
Numbers greater than 7 may result in missed deadlines for some people.

---

Grant
Darwin NT

> Einstein fix confirmed. Just update your general prefs and the next time
> it'll work. Click General pref, then Update. Don't bother changing
> anything.

Dr. Anderson figured out what was wrong after Bruce brought it to the attention of the -dev list.... the problem that was forcing you to need to update your preferences has now been solved.

---

Rob

> > In SETI I could get a cache of units for my computer to work on when the
> site
> > was down, but how does one do it in BOINC?
>
> Your Account link, View or Edit general preferences, Edit preferences.
> Under Network USage, set connect every x days to the desired number. In my
> case setting it to 6 days gives me 4 days worth of work (BOINC estimates 4.5
> hrs per Work Unit, i'm doing them in around 3hrs 15 min).
> Numbers greater than 7 may result in missed deadlines for some people.
>

Thank you Grant. I will give that a try.

Nick

---

I've noticed that the number of pending results I have has jumped up a LOT since this outage. At this point I've got something over 100 pending results across my farm.

I'm thinking that this should clear out (I expect a number of ongoing pending results, but the current number for me is several times what I was seeing earlier).

---

I've noticed that the number of pending results I have has jumped up a LOT since this outage. At this point I've got something over 100 pending results across my farm.

I'm thinking that this should clear out (I expect a number of ongoing pending results, but the current number for me is several times what I was seeing earlier).

---

> So, my question is, why do people use words like "irritate" or "bother" or
> "troublesome" when according to your message the only real effect is a bump in
> RAC?
>
Because those words reflect my reaction to problems. In particular, it represents my reaction to the longer or less accurately described outages that sometimes happens. I could ask why you question my word use, but that would suggest an incapacity on my part to understand your vantage point.

People respond to situations differently -- at least in the part of the universe I choose to occupy. Your mileage and attitudes may vary.

> Because the technology works.

I didn't say otherwise -- lots of technology works -- oft times it works less than perfectly, but it works nonetheless.

>
> I don't care that much about outages because, at the very worst, it just
> changes what I'm crunching -- and if I cared about that, I'd increase my
> cache.

Understood -- as I noted, it changed what I crunched as well -- I'm one of those somewhat disreputable sorts -- I still run Seti Classic as my alternative when BOINC is having a bad hair day (or days)

---

> I've noticed that the number of pending results I have has jumped up a LOT
> since this outage. At this point I've got something over 100 pending results
> across my farm.
>
> I'm thinking that this should clear out (I expect a number of ongoing pending
> results, but the current number for me is several times what I was seeing
> earlier).
>
>
>I just noticed that the backlog is nearing 207000 and seems to be growing by the hour. I also have about 60+ pending ones waiting to be validated and hopefully I will not lose some of them due to "old age", ie: expired time.

Best,

Nick

---

> >I just noticed that the backlog is nearing 207000 and seems to be growing
> by the hour. I also have about 60+ pending ones waiting to be validated and
> hopefully I will not lose some of them due to "old age", ie: expired time.
>
Understood -- at least one of my workstations lost work due to old age (they needed to be uploaded last Friday during the extended outage.

Speaking of outages -- at this moment as I post this we have another outage as the Berkeley folks work on the new firewall configuration which converted last weeks 6 hour outage into a 24 hour outage.

There is another thread that someone else started earlier today about that although there is no posted 'notice' of the outage.

---

Now the scheduler seems to be down...

SETI@home - 2005-05-02 22:28:35 - Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi failed

Maybe they are working to solve the validation flow problem?

---

GHz
BOINC.Italy

> Maybe they are working to solve the validation flow problem?
>

Could well be -- at this point we're working on 3 hours or so, so whatever it is, while it might be an *internally* planned outage, it amounts to an externally unannounced outage which provides an opportunity for a fair amount of speculation.

Oh well. Seti Classic is still unscathed -- I'll shift some of my farm back to it pending validated resolution.

---