| date | 2021-08-24 | ||||||
|---|---|---|---|---|---|---|---|
| title | Moving the aggregator Helm chart to public beta | ||||||
| description | The Vector aggregator Helm chart is now publicly available | ||||||
| authors |
|
||||||
| pr_numbers |
|
||||||
| release | 0.16.0 | ||||||
| hide_on_release_notes | false | ||||||
| badges |
|
We're happy to announce that our vector-aggregator Helm Chart is now
in public beta! This chart has undergone rigorous testing with our internal
design partners and is now ready for wider community feedback.
- Follow the Helm installation instructions.
- Checkout the Helm chart for a full list of options.
- Hop in our #aggregator Discord channel for help.
When Vector is deployed as an aggregator, it is used to transform and ship data collected by other agents. The core benefit of having distinct agents and aggregators is that you can have a separation of concerns within an observability data processing pipeline. Agents can become much “thinner,” in some cases acting as pure pipes that collect from a source and ship downstream to aggregators, while aggregators perform the “thicker” work of ensuring that the data is scrubbed of sensitive information, properly formatted for downstream consumers, sampled to reduce volume, and more. This is especially useful when inserting Vector into existing setups.
The vector-aggregator chart deploys Vector as a StatefulSet, with the option of
installing a HAProxy Deployment for load balancing across multiple aggregators.
Instructions on how setup and install the vector-aggregator chart can be found here.
Our next steps involve improving the out-of-the-box experience and reducing the configuration required to aggregate events from a number of common sources. We are also looking to provide a better experience integrating load balancing across multiple aggregators with additional work on the bundled HAProxy deployment.