The GitHub team released a new Go library, Vulcanizer, that interacts with an Elasticsearch cluster, yesterday. Vulcanizer is not a full-fledged Elasticsearch client. However, it is aimed at providing a high-level API to help with common tasks associated with operating an Elasticsearch cluster. These tasks include querying health status of the cluster, migrating data from nodes, updating cluster settings, and more.
GitHub makes use of Elasticsearch as the core technology behind its search services. GitHub has already released the Elastomer library for Ruby and they use Elastic library for Go by user olivere. However, the GitHub team wanted a high-level API that corresponded with the common operations on cluster such as disabling allocation or draining the shards from a node. They wanted a library that focused more on the administrative operations and that could be easily used by their existing tooling.
Since Go’s structure encourages the construction of composable software, they decided it was a good fit for Elasticsearch. This is because, Elasticsearch is very effective and helps carry out almost all the operations that can be done using its HTTP interface, and where you don’t want to write JSON manually.
Vulcanizer is great at getting nodes of a cluster, updating the max recovery cluster settings, and safely adding or removing the nodes from the exclude settings, making sure that shards don’t unexpectedly allocate onto a node. Also, Vulcanizer helps build ChatOps tooling around Elasticsearch quickly for common tasks.
GitHub team states that having all the Elasticsearch functionality in their own library, Vulcanizer, helps its internal apps to be slim and isolated.
For more information, check out the official GitHub Vulcanizer post.