Skip to content
CoRISE

Long-Term Metrics Beyond Prometheus: Thanos Retention, HA and Queries

Separate short-term collection and evaluation from long-term storage and querying.

1 min read
  • observability
  • storage
  • operations
Open table of contents

Conclusion

Separate short-term collection and evaluation from long-term storage and querying.

Context

Long-term metrics add retention, query and deduplication responsibilities. The boundary between collection and storage becomes central.

Design and verification scope

Assess the following responsibilities and boundaries when designing and verifying a configuration.

  • Prometheus retention
  • Object storage
  • Compaction
  • Queries
  • Replica labels
  • HA
  • Downsampling

Decision rationale

Use the relationship between Prometheus retention and Downsampling to compare the responsibilities of the selected approach and alternatives. Separate retained constraints from what the new boundary can change.

Trade-offs

Compare the implementation, maintenance and review work introduced by Object storage with the control it provides. Include failure paths, operator effort and conditions in which the approach should not be adopted.

Limitations

Record retention, label cardinality, deduplication, queries during store failures and cost. Topology alone does not establish performance or loss rates.

These cases provide attributed design context. They do not establish that the proposed experiments or configurations were delivered in those engagements.

Contact

Tell us about your engineering challenge.

Talk with CoRISE about the design, implementation and operation of your systems.

Start a Conversation