You are viewing a plain text version of this content. The canonical link for it is here.
Posted to test-dev@httpd.apache.org by Rodent of Unusual Size <Ke...@Golux.Com> on 2002/01/05 22:46:51 UTC
Re: cvs commit: httpd-test/perl-framework/Apache-Test/lib/Apache
TestConfig.pm
Stas Bekman wrote:
>
> How the other starting server is supposed to know whether port X is busy
> because the first server is still running or it didn't quit cleanly?
{sigh} It doesn't. But the user tries to run the tests, gets
a 'port in use' error, and runs again with '-port select'. Now
there are two servers running -- and potentially *left* running.
He gets this a few times, and codes '-port select' into his
script. Now they're propagating like rabbits because of a bug
in the harness.
> If you want to ensure that all the servers are always killed run,
>
> % killall httpd
A wonderful alternative if the test suite is the ONLY server
running. Not. And the fact that you say that's how you kill them
might explain why they sometimes get left around -- it's a bug
you don't see.
--
#ken P-)}
Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/
Author, developer, opinionist http://Apache-Server.Com/
"All right everyone! Step away from the glowing hamburger!"
Re: cvs commit: httpd-test/perl-framework/Apache-Test/lib/Apache TestConfig.pm
Posted by Stas Bekman <st...@stason.org>.
Rodent of Unusual Size wrote:
> Stas Bekman wrote:
>
>>OK, I got you. So we should make sure that the test suite will never
>>leave the server running when the testing is done.
>>
>
> Bingo.
>
>
>>Currently if for some reason this happens, harness will tell
>>you that it has failed to stop the server.
>>
>
> I am not sure that is actually the case, although your recent
> changes may have made it so.
I think at least in the cases when the server fails to start right away.
Let me know whether it's getting any better.
In any case please send the reproducable examples of the harness not
killing the server and will fix those. Thanks!
>>What are the alternative, keeping on trying to kill the server
>>indefinitely?
>>
>
> Just exit with an error status. Perhaps a spectrum of such
> exit values should be defined so a script can tell *what*
> aspect of the testing failed.
I think this happens already.
_____________________________________________________________________
Stas Bekman JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ mod_perl Guide http://perl.apache.org/guide
mailto:stas@stason.org http://ticketmaster.com http://apacheweek.com
http://singlesheaven.com http://perl.apache.org http://perlmonth.com/
Re: cvs commit: httpd-test/perl-framework/Apache-Test/lib/Apache
TestConfig.pm
Posted by Rodent of Unusual Size <Ke...@Golux.Com>.
Stas Bekman wrote:
>
> OK, I got you. So we should make sure that the test suite will never
> leave the server running when the testing is done.
Bingo.
> Currently if for some reason this happens, harness will tell
> you that it has failed to stop the server.
I am not sure that is actually the case, although your recent
changes may have made it so.
> What are the alternative, keeping on trying to kill the server
> indefinitely?
Just exit with an error status. Perhaps a spectrum of such
exit values should be defined so a script can tell *what*
aspect of the testing failed.
--
#ken P-)}
Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/
Author, developer, opinionist http://Apache-Server.Com/
"All right everyone! Step away from the glowing hamburger!"
Re: cvs commit: httpd-test/perl-framework/Apache-Test/lib/Apache TestConfig.pm
Posted by Stas Bekman <st...@stason.org>.
Rodent of Unusual Size wrote:
> Stas Bekman wrote:
>
>>How the other starting server is supposed to know whether port X is busy
>>because the first server is still running or it didn't quit cleanly?
>>
>
> {sigh} It doesn't. But the user tries to run the tests, gets
> a 'port in use' error, and runs again with '-port select'. Now
> there are two servers running -- and potentially *left* running.
> He gets this a few times, and codes '-port select' into his
> script. Now they're propagating like rabbits because of a bug
> in the harness.
>
>
>>If you want to ensure that all the servers are always killed run,
>>
>>% killall httpd
>>
>
> A wonderful alternative if the test suite is the ONLY server
> running. Not. And the fact that you say that's how you kill them
> might explain why they sometimes get left around -- it's a bug
> you don't see.
OK, I got you. So we should make sure that the test suite will never
leave the server running when the testing is done.
Currently if for some reason this happens, harness will tell you that it
has failed to stop the server. What are the alternative, keeping on
trying to kill the server indefinitely? You realize that the server may
go into uninterruptable sleep state and no kill -9 will help.
_____________________________________________________________________
Stas Bekman JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ mod_perl Guide http://perl.apache.org/guide
mailto:stas@stason.org http://ticketmaster.com http://apacheweek.com
http://singlesheaven.com http://perl.apache.org http://perlmonth.com/