PulsusDB
by digitalis.io GitHub
Technical articles from the PulsusDB team

PulsusDB
Engineering blog

Detailed articles about query compilation, telemetry data models and the ClickHouse® database capabilities used by PulsusDB

engineering.notesVOL 01
01QUERY LANGUAGESPromQL · LogQL · TraceQL
02DATA MODELSMetrics · logs · traces
03ARCHITECTUREStateless APIs · SQL pushdown
04CLICKHOUSE DATABASEExecution · storage · scale
OPEN SOURCE ENGINEERINGdigitalis.io
01 / FEATUREDLatest technical article
PLATFORM.COMPARE / 004 storageingestquery ARCHITECTURE AND TRADE-OFFS
Platform comparison12 min read

Comparing PulsusDB, SigNoz, OpenObserve and the Grafana stack

An architectural comparison of four open observability approaches. It covers storage and query execution as well as log search, deployment and day-to-day engineering work

Read the article →
02 / LATEST

More technical
Articles

Implementation details covering the storage models and query planning used for observability data in PulsusDB

02PROMQL.ENGINE / 001
PromQL

PromQL query execution with a ClickHouse® database

PulsusDB implements PromQL and the Prometheus query API over a storage model and execution path designed for large metric datasets

11 min readRead article →
03DATA.MODEL / 002
Data models

Signal-specific data models for observability telemetry

Metrics, logs and traces may share an ingestion protocol, but storing them in one generic model makes the queries people actually run more expensive

8 min readRead article →
04METRICS.ROLLUP / 003
Metrics

Raw metric and rollup selection in PulsusDB

A rollup only improves a query when its resolution and stored aggregates can preserve the meaning of the original PromQL expression

7 min readRead article →
PRODUCT REFERENCE

Platform
Documentation