Parallel runs status


Is there a way to gather the return code of the parallel runs ?

I’m willing to control the results of multiple tests ran in parallel
I want a secondary, non-parallel script, to do it

Something like that:

    - ./parallel_run:
        parallel: true
    - ./check_parallel_runs:
        parallel: false

I thought the ARTIFACT_DIR was common (I could have write the results in files) but no.


I found a way with ssh… (

In my check_parallel_runs non-parallel script I scp back all sub-nodes and do my little business then

for i in $(seq 0 $[$CIRCLE_NODE_TOTAL-1]); do 
  scp node${i}:'$CIRCLE_ARTIFACTS/*' ./
cd -

Happy to hear anyone else idea :slight_smile:

Isn’t it faster to use docker cp?
docker cp :/file/path/within/container /host/path/target

Also, I assume you have a barrier/join for all parallel runs before you copy. :slight_smile:

Sorry I see this only now.

I don’t quite get your solution. Each parallel run got its own container.