Baseline benchmark of upload and download success rate

[adamschmideg]

The idea here is to run continuously the swarm-smoke tests towards a given environment (https://swarm-gateways.net/ or a given deployment on Kubernetes), and also pushing the success or failure measurements to InfluxDB/Grafana.

So we need:

1. Cronjobs to run as part of a deployment of Swarm (https://github.com/ethereum/swarm-cluster/pull/90), for example:

• 100kb content (every 10min)?
• 1000kb content (every 1 hour)?
• 10000kb content (every 2-3 hours)?

2. Modify swarm-smoke to push success/failure measurements to InfluxDB, as well as times that it took to upload and time that it took for the test to finish. cmd/swarm: add flag for application name (swarm or swarm-private) ethereum/go-ethereum#18189

3. We should probably modify swarm-smoke to not always download concurrently from all nodes a given content upload, but to randomly pick a node and download only from it. Otherwise we are going to have a lot of the content cached on individual nodes, which might hide issues with retrieve requests.

[nonsense]

1. Cronjob is ready. swarm-private: add smoke test cron job helm-charts#13

2. We are in the process of adding configurations and flags to the smoke tests. cmd/swarm/swarm-smoke: timeout flag #1041

[nonsense]

What's left to do for this ticket is https://github.com/ethereum/swarm-cluster/issues/103 and then we can close it.

[adamschmideg]

Done