Nickcw6 commented on PR #7986:
URL: 
https://github.com/apache/incubator-devlake/pull/7986#issuecomment-2401833600

   @klesh CircleCI is functionally layered as such (where the item to the left 
contains items to the right):
   
   Pipelines -> Workflows -> Jobs -> Commands
   
   Workflows and jobs (and to a lesser extent pipelines) are what are 
statistically significant. Commands are typically small blocks of code that 
make up jobs - there is no way to collect these and little value in doing so 
anyway (not sure if these are maybe referred to as jobs in other CI tools which 
is contributing to a misunderstanding here?)  
   
   For some context & a more concrete example: a pipeline is created upon merge 
to main. Within this pipeline we often have a generic workflow called 
`deploy-to-env`. Within this will be several jobs required for the deployment, 
one of which could be called `update-prod-ecs`. 
   
   As it is this `update-prod-ecs` job that actually indicates the production 
deployment has occurred, we need to ensure it is collected and should ideally 
try to collect its finalised state as soon as possible. We want to collect the 
finished `deploy-to-env` workflow too, as this details the entire deployment 
flow.
   
   If a workflow is unfinished, then at least some of the jobs it contains will 
also be unfinished. With the current approach we'd mark both as such and 
recollect both next time. If we don't collect the unfinished jobs, then we'd 
need a mechanism to recollect these jobs based on the workflow being 
unfinished, which in my mind is more complicated than the current 
implementation. 
   
   In terms of your questions:
   
   1. Workflow timings can vary depending on what is occurring - a build 
workflow may take up to 20 minutes, whereas a deployment pipeline may be ~2 
minutes.
   2. This is more a case of the statically relevant data (ie. a successful 
deployment has occurred) being available and so we should collect it so the 
data is as up to date as possible.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to