My work computer keeps telling me that I have exceeded my daily quota and won't download any work because of that. Has there always been a daily quota or is this new? And what is the quota and how is it determined? It's kind of frustrating considering I got here a little before 8 am and found it with no workunits and a message saying it had ALREADY exceeded the quota.

> My work computer keeps telling me that I have exceeded my daily quota and
> won't download any work because of that. Has there always been a daily quota
> or is this new? And what is the quota and how is it determined? It's kind of
> frustrating considering I got here a little before 8 am and found it with no
> workunits and a message saying it had ALREADY exceeded the quota.

The max workunits setting has been around since late beta... it's 50 workunits per day. If a host tries to download more than 50 workunits in a single day, it is blocked from downloading more. This way, a host just ripping through workunits, producing crap results, can only pollute 50 results per day.

Is your cache setting really really high or something?

Rob

Nope. It is currently set to being .3-.8 days I believe. The server keeps duplicating the computer for some reason and I have had to merge it several times and each time I have to change it from home settings to work settings so that it won't download nearly a week's worth of data at a time (I found that one out the hard way, although I did process it all). Like I said before though, I got here a little before 8 am and found it had no work units and the quota was expired. I left it with maybe 3 to 5 work units left to process before it needed more work. There's not a chance it downloaded 50 work units today, or any day. The most I ever had on this computer at one time even was probably no more than 25.


>
> The max workunits setting has been around since late beta... it's 50 workunits
> per day. If a host tries to download more than 50 workunits in a single day,
> it is blocked from downloading more. This way, a host just ripping through
> workunits, producing crap results, can only pollute 50 results per day.
>
> Is your cache setting really really high or something?
>
> Rob
>
>

> Nope. It is currently set to being .3-.8 days I believe. The server keeps
> duplicating the computer for some reason and I have had to merge it several
> times and each time I have to change it from home settings to work settings so
> that it won't download nearly a week's worth of data at a time (I found that
> one out the hard way, although I did process it all). Like I said before
> though, I got here a little before 8 am and found it had no work units and the
> quota was expired. I left it with maybe 3 to 5 work units left to process
> before it needed more work. There's not a chance it downloaded 50 work units
> today, or any day. The most I ever had on this computer at one time even was
> probably no more than 25.

If your cache is not holding these work units, then your system has a problem that is causing it to generate multiple "unrecoverable errors" when processing the work. This is the reason the limit was put in place, otherwise a system can generate many thousands of these errors in a day. If you look back through your log, you will likely see that the work it got since midnight has already been returned with a client error. Under normal operations, you can see this online under the results info for that host, but this is currently disabled because of the server problems.


Darren

Yea, I was looking at my log and there seemed to be some errors in it. It's in kind of a confusing form so I'm working on putting it into a simpler form to see what happened and see if I can figure out how to fix the problem (and if this is what caused it at all). I *think* something may have happened yesterday when I detached and reattached the project. But I'm not entirely sure. The thing is set up kind of weird anyway because of BOINC's proxy bug so if there is a problem who knows where it is really coming from.


> If your cache is not holding these work units, then your system has a problem
> that is causing it to generate multiple "unrecoverable errors" when processing
> the work. This is the reason the limit was put in place, otherwise a system
> can generate many thousands of these errors in a day. If you look back
> through your log, you will likely see that the work it got since midnight has
> already been returned with a client error. Under normal operations, you can
> see this online under the results info for that host, but this is currently
> disabled because of the server problems.
>
>
> Darren
>

Well, I stuck the log in Excel and had it count the number of times that it had a certain error and it came up with 100 (50 for last night and 50 for today). It appeared it had some problem deleting a slots or something right along those lines. So I checked it out and it seems that the BOINC_CLI service had that slot locked so it didn't reset properly. I stopped/removed the service, detached the project, deleted the rogue slot directory manually along with several other configuration files including the seti project folder which didn't get deleted when I detached the project, and started the BOINC GUI back up again. This time it downloaded all of the SETI files properly and downloaded a workunit to crunch away on happily.

So, that basically leaves me with another question. Does running the GUI cause a conflict with the CLI, or did it just happen because I forgot to kill the BOINC service before I detached/reattached yesterday to update the server (which I'm not sure if I even installed the service at that point)?


> Yea, I was looking at my log and there seemed to be some errors in it. It's in
> kind of a confusing form so I'm working on putting it into a simpler form to
> see what happened and see if I can figure out how to fix the problem (and if
> this is what caused it at all). I *think* something may have happened
> yesterday when I detached and reattached the project. But I'm not entirely
> sure. The thing is set up kind of weird anyway because of BOINC's proxy bug so
> if there is a problem who knows where it is really coming from.
>
>

> So, that basically leaves me with another question. Does running the GUI cause
> a conflict with the CLI
Yes. The schedule has the GUI controlling the CLI which controls the projects and the connections. But this is not there yet.

I'm not quite sure I follow what you are saying. What do you mean "this" is not there yet? The CLI is there. The GUI is there. So are you saying that the GUI is SUPPOSED to control the CLI and doesn't?

> > So, that basically leaves me with another question. Does running the GUI
> cause
> > a conflict with the CLI
> Yes. The schedule has the GUI controlling the CLI which controls the projects
> and the connections. But this is not there yet.
>
>

> I'm not quite sure I follow what you are saying. What do you mean "this" is
> not there yet? The CLI is there. The GUI is there. So are you saying that the
> GUI is SUPPOSED to control the CLI and doesn't?
>
> > > So, that basically leaves me with another question. Does running the
> GUI
> > cause
> > > a conflict with the CLI
> > Yes. The schedule has the GUI controlling the CLI which controls the
> projects
> > and the connections. But this is not there yet.
> >
> >
The the windows CLI has gotten very little testing so it is not really ready for use. There is supposed to be a new GUI created that will control the CLI, at that time the current GUI will disappear.

John Keck
BOINCing since 2002/12/08