Hey folks, we pushed 2.0.0-beta3
this morning
There’ve been two main focuses in this release: improving XTDB’s interoperability with language-standard Postgres tooling, and improving the deployment/operational story.
- There’s now an Azure-specific Kubernetes guide, but this should be broadly applicable to wherever you get your k8s from.
- We’ve added simple
/healthz
endpoints for you to point your healthchecks at - addhealthz: {port: 8080}
/:healthz {:port 8080}
to your YAML/EDN config respectively. - Under the hood, we’ve made significant changes to the buffer pool to allow us to more accurately account for the memory/disk we’re using - these should be available as monitoring metrics shortly.
- Following some recent changes to the compactor, we’ve been able to significantly reduce the work necessary in the indexer single-thread (the compactor runs in the background and is highly parallelisable) - this has improved the time-to-query-availability for a bulk load by ~30-35%.
- For more details, including some minor breaking changes, see the release notes
As well as the Clojure API, you can try it out with Postgres tooling:
docker run --rm --name xtdb -ti -p 5432:5432 ghcr.io/xtdb/xtdb:2.0.0-beta3
psql -U xtdb -h localhost
(for the Clojure folks)
(require '[next.jdbc :as jdbc])
(with-open [conn (jdbc/get-connection "jdbc:postgresql://localhost/xtdb")]
(jdbc/execute! conn ["INSERT INTO foo RECORDS {_id: 'foo', ...}"]))
As always, any questions or issues, give us a shout
James & the XT Team