InfluxDB Cloud Dedicated v2.12.0 introduces a new `--perf-debug` flag for the `influxctl` query command that provides real-time query performance metrics without returning actual results. The feature displays execution metrics in both human-readable tabular and JSON formats, including client duration, compute duration, memory usage, parquet files accessed, and ingester metrics. This enables faster debugging and optimization cycles by eliminating the need to correlate queries with system logs. The post explains key metrics like cache warmup effects, data density impacts, and planning versus execution duration, along with practical troubleshooting tips for identifying performance bottlenecks in time-series queries.

8m read timeFrom influxdata.com
Post cover image
Table of contents
Sample outputNotesPotential sources of latency or variabilityOther things to try
2.2K Impressions