Has anyone else had this error:

ACTIVE_TASK_SET::insert() can't delete file slots\\\stderr.txt
Unrecoverable error for result (Couldn't start the app for this for this result: error -110)

I had around 20 WU (lucky me) then had a series of these errors and they all got zapped! :(

It's download a couple of WUs since and seems to be running alright since...

Mark

First. Shut down BOINC. That will keep it from doing any more. The only thing you will be able to do is download lots of workunits only to have them error like that and not process. You can only go through 50 a day, but it's still a big waste (I went through 100).

Second, I'm not exactly sure what caused this in your case, but for me, it was related to running the CLI and the GUI at the same time (being unaware that the GUI didn't work with the CLI properly yet). What happened was the CLI locked files and the GUI couldn't delete/overwrite them with new workunit data for processing. If you have the CLI running, you will need to stop the CLI. BOINC *may* work at that point, but I can't say for sure. When I figured out that the CLI was causing the problem because it had locked files, I closed it and the GUI down, uninstalled it as a service, and then went to the "slots" folder in the BOINC directory, and deleted everything there, and then restarted the GUI.

Since I'd already downloaded my 50 quota for the day, it wouldn't let me download anymore units until the next day. I didn't really want to wait that long, so I *think* what I did was delete the scheduling server line in one of the XML settings files and was able to download again. From reading some of the Alpha testing messages, it seems there is a setting in the file that records the number of downloaded workunits for the day which you could probably change to get it to download once you get your problem worked out. I don't remember what that setting is, but I'm sure someone else around here could provide a little more specific and definite information about what to change/delete to get BOINC to download again (without losing any workunits already downloaded). If you don't have any workunits downloaded, you could probably just reset the project or detach/reattach, but again, I've only had to fix the problem once, and unfortunately I don't have time to do more research on it. I hope this helps though.

Chris



> Has anyone else had this error:
>
> ACTIVE_TASK_SET::insert() can't delete file slots\stderr.txt
> Unrecoverable error for result (Couldn't start the app for this for this
> result: error -110)
>
> I had around 20 WU (lucky me) then had a series of these errors and they all
> got zapped! :(
>
> It's download a couple of WUs since and seems to be running alright since...
>
> Mark
>

Thanks for that.

Actually, since shutting it down and restarting from zero WUs it seems ok now. It's gone through a complete cycle (download, process, upload) without any problems. I'll keep an eye on it just in case.

I don't know if this is related or not. Yesterday, while there was still a backlog of results to upload (I had nearly 100 on this PC), that the messages tab seemed to have some sort of problem (resource?). With all the errors from the failed uploads the log quickly filled up so that changing tab causes a major delay (of minutes) whilst the screen was out of date. Maybe they should flush out the old messages every so often...

Mark

> First. Shut down BOINC. That will keep it from doing any more. The only thing
> you will be able to do is download lots of workunits only to have them error
> like that and not process. You can only go through 50 a day, but it's still a
> big waste (I went through 100).
>
> Second, I'm not exactly sure what caused this in your case, but for me, it was
> related to running the CLI and the GUI at the same time (being unaware that
> the GUI didn't work with the CLI properly yet). What happened was the CLI
> locked files and the GUI couldn't delete/overwrite them with new workunit data
> for processing. If you have the CLI running, you will need to stop the CLI.
> BOINC *may* work at that point, but I can't say for sure. When I figured out
> that the CLI was causing the problem because it had locked files, I closed it
> and the GUI down, uninstalled it as a service, and then went to the "slots"
> folder in the BOINC directory, and deleted everything there, and then
> restarted the GUI.
>
> Since I'd already downloaded my 50 quota for the day, it wouldn't let me
> download anymore units until the next day. I didn't really want to wait that
> long, so I *think* what I did was delete the scheduling server line in one of
> the XML settings files and was able to download again. From reading some of
> the Alpha testing messages, it seems there is a setting in the file that
> records the number of downloaded workunits for the day which you could
> probably change to get it to download once you get your problem worked out. I
> don't remember what that setting is, but I'm sure someone else around here
> could provide a little more specific and definite information about
> what to change/delete to get BOINC to download again (without losing any
> workunits already downloaded). If you don't have any workunits downloaded, you
> could probably just reset the project or detach/reattach, but again, I've only
> had to fix the problem once, and unfortunately I don't have time to do more
> research on it. I hope this helps though.
>
> Chris

A quick way to flush the old messages is to exit the client, then restart.


Alpha Tester

---

Account frozen...