Aura is an in-memory time-series data store with the ability to flush "segments" to other long-term storage systems.
Taking inspiration from Facebook's Gorilla/Beringei and Pinterest's Yuvi in-memory stores, Aura compresses time-series data following the Gorilla format with the added capability to perform additional lossy compression to save space.
The store is meant to hold the most recent time series data with a query engine like OpenTSDB 3.x routing queries to longer term storage for older data.
Configuration is performed via a ShardConfig and OpenTSDB configuration class.
Future work will merge them.
Currently, the repo contains the individual libraries for Aura components. We'll will tie in OpenTSDB 3.0's write path with Kafka and Pulsar streaming input as well as the OpenTSDB query servlets.
To store data a TimeSeriesStorageIf instance must be instantiated with a
ShardConfig. Data can then be written via
TimeSeriesShardIF.addEvent(HashedLowLevelMetricData data) where the data is
an implementation of the low-level interface.
Data can then be read out using the AuraMetricsSourceFactory and OpenTSDB query
layer.
Work is ongoing to integrate flushes and reads with an Aeropsike cluster for long-term retention.
Please see the Contributing file for information on how to get involved. We welcome issues, questions, and pull requests.
- Smruti Ranjan Sahoo @smrutilal2
- Arun Gupta @arungupta
- Ravi Kiran Chiruvolu @mailravi3390
- Chaitanya GSK @gskchaitanya
- Chris Larsen @manolama
This project is licensed under the terms of the Apache 2.0 open source license. Please refer to LICENSE for the full terms.