Re: Why does Boinc say "Not reporting or requesting tasks"?

Tazz <dmswaine@gmail.com> wrote:

<snip>

Anyhow, that's about all I know about it. I have two other questions.
One is, what does "Connected to localhost (6.6.31)" at the bottom of
the Boinc screen mean? It is not a dotted decimal IP address or a
port number. The other is, is it possible to run Seti on other
computers and obtain work from a central Boinc? I used to do this
with SetiGate; I don't know if this is possible with Boinc. (Among 
other things, I was thinking about depleting the queue when it gets
near one of the windows, to trigger fetching of work.)

Thanks for any info. Yours, Dave


The connected to localhost message is to tell you what computer BOINC 
manager is monitoring. If you had a network of computers all running 
BOINC you could use the BOINC manager from one computer to check on all 
of the others one at a time. From BOINC go to Advanced> Select 
computer... But first you have to set up the remote_hosts.cfg and 
gui_rpc_auth.cfg files on the other computers (the ones you want to 
monitor).

The 6.6.31 is the BOINC version number for the computer you're 
monitoring or the version number of the BOINC manager you're using, I 
can't remember which. If you were monitoring a different computer the 
message would say "Connected to <computername> (BOINC version number)".

Hmm. This would seem to imply that each computer must necessarily have its own Boinc and Boinc
queue, since they may be running different OS'es, etc., yes? No central Boinc repository, eh.

Btw, still in "not requesting new tasks" mode. All failed uploads have cleared. All "ready to
reports" have cleared. Queue contains 24 tasks including 3 running. Have downloaded just enough work
to replace those that have completed. Most are long "To completion" times > 10 hrs.

I just changed the "additional work buffer" to 3 days (rather than 2.5) and Boinc immediately went
online and started downloading work, 4 units, all > 10 hours.

This leads to several hypotheses which, if correct, would more or less explain my original question
regarding the Boinc downloading behavior. 1) Requesting work from SAH is based on the number of
hours "To completion" in the queue, not the number of tasks. This is as it should be, and would
explain why the queue "runs out" when a lot of tasks run much shorter than expected. (There are
those who would say this should be obvious, but it may not be so obvious when the queue bloats up
with 160-200 tasks, some with very short deadlines.) 2) Tasks in the state, "running", "uploading",
"upload pending", "suspended", "ready to report", etc. count towards this hourly total. This would
explain, for example, why a bunch of failed uploads prevents further fetching. 3) "Downloading" and
"Download Pending" don't count towards this total. This would explain why, when a bunch of
"Downloading's" fail and switch to "Download Pending", it triggers even more downloads, causing
possible bloating. Oh, and 4) It appears possible that numbers to the right of the decimal point in
the "additional work buffer" field in the preferences have no effect; I haven't really tested this
out.

It might be useful to move the "additional work buffer" setting out of the "preferences" menu and
put it somewhere prominent, for example, on the "projects" or "tasks" panel, express it in hours or
days:hours, since that seems to be the way Boinc calculates it, and break it down into requested and
actual for Seti, cuda, and Astropulse so that it would be possible to keep a handle on the flow of
work. Just my 2�.