How do metrics get into Prometheus? How does it differ from InfluxDB (push vs pull model)?

Prometheus uses a pull model for metric collection: it periodically polls (scrapes) HTTP endpoints of applications or services that expose metrics in a format understandable by Prometheus. This allows Prometheus to control the collection frequency and simplifies the discovery of new targets.

InfluxDB, on the other hand, is more often used with a push model, where agents or applications send (push) metrics to the database themselves. This is convenient for systems where polling is difficult or impossible to organize, such as in short-lived tasks.

Main differences:

• Pull (Prometheus):

  • Prometheus initiates requests to targets.
  • Simplifies monitoring of dynamic environments.
  • Requires targets to be accessible over the network.

• Push (InfluxDB):

  • Clients send data themselves.
  • Suitable for isolated or short-term services.
  • Requires configuration of agents or clients for sending.

Thus, the choice of model depends on the system architecture and monitoring requirements.