Closed tpaz closed 6 years ago
I replaced "run:" with "local:" command and it seems to work now.
So bug is on README. need to fix the above in "Running Sup from Supfile" section.
This seems to be fixed for a while..
It's not fixed on the website. I assume the HTML is somehow generated from the readme. If not, I can create a PR.
Sup version: 0.5 (pulled today)
Following the README documentation, I tried to use sup within Supfile. Either there is a bug or the documentation is misleading. Instead of instantiating a sub-process as the README says on the machine where sup is executed, it actually tries to run the sup tool on the Remote machine where sup (parent) is supposed to execute the "run" command.
example: [results@control experiment]$ sup worker1 cloud_suite_bench root@worker1:22 | bash: sup: command not found root@worker1:22 | Process exited with status 127
Supfile code: cloud_suite_bench: desc: install && run cloud_suite benchmark (based on ENV vars) run: > sup -f ./cloudsuite/Supfile $SUP_ENV $SUP_NETWORK run_bench
The "sub" sup file is in ./cloudsuite folder as hinted above, here is its content just for reference:
commands: run_benchmark: desc: install and run cloud_suite benchmark. Prerequisites - Docker + Compose + swarm cluster configured on 3 worker nodes. upload:
TODO: collect results
cleansing: desc: clean all leftovers after bench over. just to make sure... run: > docker-compose down targets: run_bench:
I suppose it has nothing to do with the 'sub' sup file, rather the fact that the "run:" command is expected to execute its arguments on the remote worker machine rather on the control machine where it was initiated. This is in contrast to the expected behavior based on the README description.
Ideas? TIA