Hi all,

So I have the following requirements for the CI of our big monolith and I'd 
like to know if and how this is best achieved with Jenkins.

- We have many different machines that we want to utilize as workers
- Each machine will host many workers (as many as its cores, actually)
- A single run of our test suite should be parallelized among different 
different workers (our project supports this). The command each worker will 
use is something like this: `./tests.sh --build foo --worker bar` and this 
will be spawned via a docker-compose.yml
- I want to have many "build queues" for my project, so that when a user 
executes a build, another one is not blocked waiting for the first one. So 
each worker must somehow be assigned to a queue (labels?)
- I want to be able to use Pipelines declared inside the repo

How would you go about this? Should I have a "Jenkins agent per worker"? Do 
I have to have a "node per machine (worker host)"? What about # of 
executors? How about the pipeline itself: can I tell it somehow to run 50 
instances of that same command (1 in each worker)?

Thanks in advance.

-- 
You received this message because you are subscribed to the Google Groups 
"Jenkins Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jenkinsci-users/d20daf0b-ce15-44f0-b54c-b0bf6b96ad82%40googlegroups.com.

Reply via email to