Prometheus query to determine if a MonoVertex has healthy data movement

[juliev0]

Summary

Assuming Numaplane uses Prometheus querying, we need a query to determine if a MonoVertex has healthy data movement. It should return "healthy" state if there's no data flowing into the source.

---

Message from the maintainers:

If you wish to see this enhancement implemented please add a 👍 reaction to this issue! We often sort issues this way to know what to prioritize.

[juliev0]

Received query from @whynowy:

increase(numaflow_monovtx_ack_total{namespace="oss-analytics-sampledataflow-usw2-prd", mvtx_name="numa-endurance", mvtx_replica="0"}[1m])
(check this for > 0)

so I will close this

[whynowy]

Received query from @whynowy:

increase(numaflow_monovtx_ack_total{namespace="oss-analytics-sampledataflow-usw2-prd", mvtx_name="numa-endurance", mvtx_replica="0"}[1m]) (check this for > 0)

so I will close this

Be aware this is just an example query, not the final one.

[juliev0]

Received query from @whynowy:
increase(numaflow_monovtx_ack_total{namespace="oss-analytics-sampledataflow-usw2-prd", mvtx_name="numa-endurance", mvtx_replica="0"}[1m]) (check this for > 0)
so I will close this

Be aware this is just an example query, not the final one.

ahh, got it, thanks - in that case I will reopen this

[whynowy]

Received query from @whynowy:
increase(numaflow_monovtx_ack_total{namespace="oss-analytics-sampledataflow-usw2-prd", mvtx_name="numa-endurance", mvtx_replica="0"}[1m]) (check this for > 0)
so I will close this

Be aware this is just an example query, not the final one.

ahh, got it, thanks - in that case I will reopen this

@juliev0 - Be aware those queries should be configurable in Numaplane, so it really doesn't matter what they are, as long as all the queries are executed and the conditions are met, Numaplane can consider it's successful.

[juliev0]

Received query from @whynowy:
increase(numaflow_monovtx_ack_total{namespace="oss-analytics-sampledataflow-usw2-prd", mvtx_name="numa-endurance", mvtx_replica="0"}[1m]) (check this for > 0)
so I will close this

Be aware this is just an example query, not the final one.

ahh, got it, thanks - in that case I will reopen this

@juliev0 - Be aware those queries should be configurable in Numaplane, so it really doesn't matter what they are, as long as all the queries are executed and the conditions are met, Numaplane can consider it's successful.

Yes, of course it's an abstraction. But if we're going to do end to end testing, we need for it to work as well of course. As long as the query we have only runs the risk of false negatives and not much false positive (for detecting errors), then that meets our need. In any case, we have some time.