Re: disk space limit of scallywag ?

[Marco Atzeri <marco.atzeri@gmail.com>]

On 12.03.2023 22:00, Jon Turney wrote:
> On 12/03/2023 20:33, Marco Atzeri via Cygwin-apps wrote:
>> On 12.03.2023 18:39, Jon Turney wrote:
>>> On 12/03/2023 17:08, Marco Atzeri via Cygwin-apps wrote:
>>>> Hi Jon,
>>>>
>>>> what is the current disk space limit of scallywag ?
>>>>
>>>> jobs 5594
>>>>
>>>> https://github.com/cygwin/scallywag/actions/runs/4397354263/jobs/7700519955
>>>>
>>>> ..
>>>> scallywag: staging/cmake/vim-cmake/vim-cmake-3.23.2-2.hint
>>>> scallywag: staging/cmake/vim-cmake/vim-cmake-3.23.2-2.tar.xz
>>>> xz: (stdout): Write error: No space left on device
>>>> tar: /cygdrive/d/a/scallywag/scallywag/builddir.tar.xz: Cannot 
>>>> write: Broken pipe
>>>> tar: Child returned status 1
>>>> tar: Error is not recoverable: exiting now
>>>> Traceback (most recent call last):
>>>>
>>>
>>> The the VM is supposed to have at least 14GB free disk space, which 
>>> seems like it should be enough.
>>>
>>> But maybe there's something I'm doing wrong here (like trying to 
>>> create this archive on the wrong drive or something...).  Let me look 
>>> into this a bit more...
>>>
>>

>>
>> just the testing is requiring
>>    3.6G    Tests
>>
>> so no doubt it hits the 14GB issues
> 
> Yeah. It seems to be the test suite which takes a fantastic amount of 
> disk space (and time), so as a workaround for the moment, adding the 
> control token 'notest' should let it build and package successfully.
> 

Hi Jon,
will be possible to pass the token notest to the sub-command rerun ?
For the time being I used SCALLYWEG="notest"

PS: there is a mismatch in the documentation

subcommands:
   {cancel,deploy,help,rerun}
     cancel              cancel job
     deploy              deploy job
     help                this help
     rerun               re-run job

While on the web:

rebuild
rebuild a job (e.g. if it failed due to some transient condition, or 
optionally with different tokens).


Regards
Marco