v0.4.0

Note

In previous releases, a json array like

{
	"name": "John",
	"age": 30,
	"cars": [{
			"brand": "Ford"
		},
		{
			"brand": "BMW",
			"model": "x1"
		},
		{
			"brand": "Audi",
			"model": "q1"
		}
	]
}

would be flattened to cars.0_brand,cars.0_model,
cars.1_brand, cars.1_model, cars.2_brand,
cars.2_model. This was an issue for very long
arrays, a single event with a long array could add
several columns to the schema. But most of the times these
columns would be empty.

This release changes the approach to flatten to two fields
like cars_brand, cars_model. Each column will have
an array of elements in that field.

Refer #344

What's Changed

• Use delta encoding for timestamp by @trueleo in #341
• Add benchmark details by @nitisht in #342
• Refactor handlers by @trueleo in #343
• Fix schema deserialization in case of missing metadata by @trueleo in #345
• Bump openssl from 0.10.45 to 0.10.48 by @dependabot in #346
• Add user analytics by @nitisht in #347
• Alert log by @nitisht in #348
• Create CITATION.cff by @nitisht in #351
• Update CITATION.cff by @nitisht in #352
• Add support for regex based rule in alerts by @nitisht in #353
• Add vector and fluent-bit subcharts by @nitisht in #355
• Add custom flattening that ignores array by @trueleo in #344
• Add securityContext for Pod to fix PV mount permissions by @nitisht in #356
• Prepare for next release v0.4.0 by @nitisht in #357

New Contributors

• @dependabot made their first contribution in #346

Full Changelog: v0.3.0...v0.4.0