Red Hat Bugzilla – Bug 166225
no Job ID was provided
Last modified: 2007-04-18 13:30:37 EDT
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.10) Gecko/20050719 Fedora/1.7.10-1.5.1
Description of problem:
# make plague
/usr/bin/plague-client build perl-Sort-Versions perl-Sort-Versions-1_5-3_fc4 fc4
Package perl-Sort-Versions enqueued. (However, no Job ID was provided in the time required)
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. make plague
Actual Results: see above.
this just happened to me:
/usr/bin/plague-client build perl-XML-LibXSLT perl-XML-LibXSLT-1_58-1_fc5 devel
Package perl-XML-LibXSLT enqueued. (However, no Job ID was provided in the time
I waited about an hour then tried again and the package was successfully
enqueued & returned a job id.
If the server is under heavy load at the time, or takes too long to enqueue the
job (currently 3 seconds) it won't pass the ID back to the client. Its still
queued and you can find the job in the Web interface though. This won't really
be fixed without larger changes to the codebase, so marking WONTFIX as it
doesn't really affect functionality, just presentation.
(In reply to comment #2)
> If the server is under heavy load at the time, or takes too long to enqueue the
> job (currently 3 seconds) it won't pass the ID back to the client.
I guess, are you referring to times in a local network?
> Its still
> queued and you can find the job in the Web interface though.
What you say doesn't match with my observations: At the time this had happened,
the job could NOT be found with "plague-client list".
Actually, I had invoked "make plague" ca. 10 times, but not a single job had
been queued. It's just just that invoking "make plague" a new several hours
laters, had succeeded.
Also see the detailed experience report in comment 1 and notice that
there's only a single job for perl-XML-LibXSLT-1_58-1_fc5 listed in
the web interface (job 126). So the build system did not enqueue the
first request, but only the second one an hour later.
Reopening, because comment #2 apparently is not correct.
Wrt. to timeouts, I am on a DSL line, being connected to the buildsystem via a
transatlantic connection. I.e. if comment #2 should refer to internet
connections, 3 second timeouts probably is at least one magnitude too small.
Just to confirm, comment #4 is correct, i didn't see my job listed in the web
interface that first time, only the second time.
I'm on a cable modem on the east coast
Closing, this is a transient error and should usually be fixed with a
buildsystem kick when the build system is acting up.