Posts by =Lupus=

1) Message boards : Closed Issues : HTTP error 413, client intended to send too large body (Message 3047)
Posted 17 Feb 2019 by =Lupus=
Post:
Worked.

Seems my client wanted to report the first 100 to 200 failures.
Which your webserver did not like.
And when i killed the already-closed workunits i generated an even bigger number…

1 Megabyte is way too much for a "normal" request but, well, it is BOINC.
All report-completes and everything else goes thru normal HTTP requests.
Which is easy because the needed protocols are on every OS but seems to need a Little bit tweaking when handling big requests or reportings.

Happy I could help you find that "error" in early-alpha phase instead of later on when hundreds of users are there...
2) Message boards : Closed Issues : HTTP error 413, client intended to send too large body (Message 3041)
Posted 17 Feb 2019 by =Lupus=
Post:
Getting Error 413 on reporting. Not your executable but a BOINC problem.

17.02.2019 19:50:06 | Tomáš Brada Experiment Grid | Sending scheduler request: Requested by project.
17.02.2019 19:50:06 | Tomáš Brada Experiment Grid | Reporting 968 completed tasks
17.02.2019 19:50:06 | Tomáš Brada Experiment Grid | Not requesting tasks: "no new tasks" requested via Manager
17.02.2019 19:50:08 | Tomáš Brada Experiment Grid | Scheduler request failed: Error 413


Retried manually, persists the retries.
3) Message boards : Closed Issues : STATUS_ACCESS_VIOLATION, padls rows 7 an 8 (Message 3040)
Posted 17 Feb 2019 by =Lupus=
Post:
HTTP error 413 still there.
So I dont really get the updated info from server side.

Manually aborted the tasks as all of them were of the 7.... batch which got cancelled.

Seems error 413 means "answer too big" - either my or your side cant handle the size of the request.
Something the BOINC makers should know about.
Will open new thread.
4) Message boards : Closed Issues : STATUS_ACCESS_VIOLATION, padls rows 7 an 8 (Message 3038)
Posted 17 Feb 2019 by =Lupus=
Post:
Hello,

as I read about "cancelling WUs with 8 or 7 as first Digit in start row - is it OK for me to cancel WUs which are named "padls2_7..." as they have a leading 7 in start row?
They seem to be killing themselves in seconds or running for Hours (6+ Hours Right now)

Or are the non-self-destruction WUs ok and they run that long?

Greetings,

=Lupus=

*EDIT* nevermind, just saw all WUs went into Status "not needed anymore" - but my Client does not update because BOINC Scheduler sends me http error 413.




©2020 Tomáš Brada