timbala
timbala copied to clipboard
Durable time-series database that's API-compatible with Prometheus.
Using [Jumphash](https://arxiv.org/abs/1406.2294), nodes are represented by buckets. In order to minimise the amount of data that moves when nodes are added or removed, nodes must remember which bucket they belong...
When determining which nodes samples should be forwarded to, avoid duplicating the same set of samples in memory *n* times, where *n* is the replication factor: https://github.com/mattbostock/timbala/blob/f0109697dbbf547f118d9e8b17a5852c979ca624/internal/write/write.go#L122-L137
Design a logo for the project.
Allow reads and writes to be traced, to help debug performance issues, functional issues, or bottlenecks. The [OpenTracing](http://opentracing.io/) standard should be used. Potentially useful docs: [OpenTracing specification](https://github.com/opentracing/specification/blob/master/specification.md) [OpenTracing semantic conventions](https://github.com/opentracing/specification/blob/master/semantic_conventions.md)...
Like #54 but for incoming queries. Any node in the cluster should be able to respond to queries and retrieve data from the peers responsible for the time-series being queried.