We’ve identified several opportunities across our key open-source projects – Telegraf, InfluxDB, and Kapacitor. Below, we’ve laid out the details for each.
It’s been said that a storage engine is never really done, so our TSM engine is always seeking improvement. Potential projects include performance optimization, profiling on different hardware, and expanding test coverage.
InfluxDB’s iterator-based query engine is a relatively new design, so tuning the design for different query patterns is always a possibility. Additionally, there are many new aggregators and transformations that could be added to add functionality in other fields, such as statistical analysis.
Optimize the line protocol parser and reduce unnecessary allocations.
Improved system stats
With some help, Telegraf (and possible also gopsutil) could be better at getting system metrics. Some areas that need particular help are: netstat (currently using lsof) and Windows perf counters over WMI for system info on Windows.
Better automated testing
General help with getting a better automated testing system together, integration testing, and testing multiple outputs and inputs on multiple OSes.
Telegraf can always benefit from help with writing plugins. Current interests are: gathering AWS metrics, tail log file parsing (with templates for common logs), see our open issues for others.