[Piuparts-devel] Bug#526045: Bug#526045: Patch

Scott Schaefer scott at neurodiverse.org
Sun Jun 26 19:43:09 UTC 2011


On 06/25/2011 06:01 PM, Holger Levsen wrote:
> welcome back! :-)
>
> On Freitag, 24. Juni 2011, Scott Schaefer wrote:
>    
>> Ping  (I'm back).  I should be able to respond relatively quickly to any
>> questions/issues over next 8-10 days.
>>      
> I've applied your patch now to trunk and am running it on piatti - in a few
> hours results will be visible on piuparts.d.o
>
> i'm looking much forward to it! :-)
>
> dump_pkgs() is not used at all atm - i was also thinking of some "default way"
> to query state and such....
>
>
> cheers,
> 	Holger
>    

My $0.02 on dump_pkgs() ...

This was a first-cut attempt toward improving ability to test mods to 
piuparts-master and/or piuparts-slave.

I believe the most useful solution here is to code a signal handler 
(SIGUSR1, SIGUSR2) to cause the master to dump state of packages to a 
file.  I have "research Python signal handlers" on my TODO list.

Then, I want to add support in the slave to allow for "stub testing"; 
e.g. the slave would not actually test the package, but would 
immediately report "pass" for all packages except those whose names are 
found in (command arg specified) input file.  Something like:

package-A = fail
package-B = fail @logfile-1
package-C = fail @logfile-2

I'm treating both of these as "wishlist" for now.








More information about the Piuparts-devel mailing list