Open craig-landry opened 3 years ago
To flesh this out some more, the ultimate goal is to get to a number of caring actions that cht-core 4.0.0 can support per month documenting things like number and size of servers required so we can give partners a good idea of how many users they can scale to given a known workload per user.
Obviously we don't want to run the suite for a whole month, so the script will need to extrapolate from a smaller scale run.
@lorerod has made a great start on this script for generating production like data here: https://github.com/medic/cht-core/tree/master/tests/scalability/generate_brac_data . The rest of the work is to write a script which uploads the generated data and gets a bunch of users to do the initial replication, then simulating cht-core by replicating bidirectionally periodically to upload and download docs.
@dianabarsan has agreed to take this work on when convenient so assigning over to her.
Output should be reliable, realistic metrics of what we would expect for CHT 4.0/archv3 in production against a large CHT instance.