v1.1.15

1. Use zlib compression
2. Add statsd metrics for keys added and keys deleted
3. Improved compaction and key deletion latency
4. Expose badger options as follows:

badger: 
  levelOneSize: 204800000
  maxLevels: 3
  syncWrites: false

v1.1.14

• Added compaction prior to server shutdown

v1.1.13

• Fix the compaction for the key whose hash key value is different from the last merged key.
• Change the compression to snappy for faster writes.
• Add stats for error in ingestion and compaction

v1.1.12

• Added an ability to write to multiple sinks
• Added net module for LUA exposing a way to query an IP/MAC addresses

v1.1.11

• Added liveness & readiness probe server endpoint for envoy
• Fixed BigQuery writer which defaulted to CSV instead of ORC
• Fixed timestamp column in ORC
• Changed make://timestamp to output the value as a timestamp instead of int64

v1.1.10

• Fixed an issue where computed columns were not computed for all rows under very load.
• Changed the docker image workflows. The edge tag is now built only for PRs, latest for every master and :v1.x.x for every release.

v1.1.9

• A unique event identifier can be generated by using make://identifier function in a computed column.
• An ingestion time stamp can be generated by using make://timestamp function in a computed column.

In order to use it, simply declare a computed column as such:

computed:
  - name: ingested_at
    func: "make://timestamp"
  - name: id
    func: "make://identifier"

v1.1.8

• Added better initialization logs

v1.1.7

• Added support for loading computed columns and configs from Google Cloud Storage. This can be done by using gs://bucket/prefix scheme when specifying the url.

v1.1.6

• Added ingestion by CSV and URL (for both CSV and ORC files)
• Added certs to the docker image