Recommendations to others considering Google Cloud Observability:
Useful for aggregating and searching logs, but less useful for metrics. Other products like Datadog have better support for being able to easily send metrics from each process. With Stackdriver metrics, you need to aggregate them yourself because you can only send one metric every 5 seconds or so. If you have two microservice instances, that means sending the metrics to a dedicated microservice just to collect them and aggregate them so that between your entire system, the metric is only sent to GCP every 5 seconds or so. Review collected by and hosted on G2.com.
What problems is Google Cloud Observability solving and how is that benefiting you?
Knowing what's going on in each stage of our big data pipeline. Logs from different microservices are aggregated and we can search through them. It supports structured logging, so we can attach tags to the log messages like the Kubernetes pod or cluster it's running in, to more easily filter the messages to what we want to see. Review collected by and hosted on G2.com.