OnGres Inc./StackGres: Release 1.5.0

Release: 1.5.0

Released: 2023-07-04

License: AGPL-3.0

Release Assets:

# :rocket: StackGres 1.5.0 GA (2023-07-03) The Postgres You Love :heart_eyes:, At Any Scale :sparkles: We are thrilled to announce the release of StackGres 1.5.0! This update brings a host of new features, enhancements, and optimizations, further solidifying StackGres as the go-to solution for running your mission-critical Postgres workloads on Kubernetes. Let's dive into the exciting highlights of this release: 1. :fire: Sharding Cluster Support: StackGres 1.5.0 introduces seamless integration with Citus and Patroni, empowering you to scale your Postgres database horizontally. Now you can effortlessly shard your clusters, distributing data across multiple nodes for improved performance, scalability, and fault tolerance. Citus uses a technique called sharding to divide your data into smaller, manageable pieces called shards. Each shard contains a subset of your data and is stored on a separate node within the cluster. Sharding allows for parallel processing of queries and efficient data storage across multiple machines. When you execute a query, Citus intelligently routes the query to the relevant shards. It analyzes the query and determines which shards hold the necessary data to process the request. By distributing the workload across multiple nodes, Citus enables parallel query execution, leading to faster query performance. Citus provides a scalable and distributed database solution by leveraging PostgreSQL's powerful extensibility. It enables you to handle large datasets, process complex queries, and achieve high performance by harnessing the capabilities of a distributed cluster architecture, as you can add more nodes to your cluster to accommodate increasing data volumes or user traffic. A new CRD called `SGShardedCluster` will bring you the power of sharding by creating multiple SGClusters that will behave like a single one. This also features an automated SSL/TLS support by default on sharded clusters. StackGres provides with the `SGShardedCluster` a fault tolerance HA solution on each shard (as well as the coordinators) so that any failed node will be automatically recovered as if one node goes down, the other nodes can still continue serving your application. The Web Console has also been updated with first-class support for this feature, try it out :wink:. | :warning: WARNING | |:---------------------------| | The feature is still in alpha so you might expect breaking changes in the CRD, if you encounter any issue please report it and we will provide a fix. Also, 2 day operations are currently not supported and will be implemented in 1.6 release. | 2. :elephant: Updated Postgres and components: In this release, we have updated the Postgres engine to the latest minor versions supported up to 15.3, 14.8, and 13.11, incorporating the latest advancements and bug fixes from the PostgreSQL community. This upgrade brings improved performance, enhanced security features, and better compatibility with the latest Postgres ecosystem. Also, we have updated Babelfish for PostgreSQL to the latest release 2.3.0 so you can continue experimenting with this SQL Server alternative with expanded language support, improved compatibility and bug fixes and stability improvements. 3. :anchor: Compatibility with Kubernetes 1.27 and 1.26: StackGres is now fully compatible with Kubernetes versions 1.27 and 1.26. This means you can confidently deploy and manage StackGres in Kubernetes clusters running these specific versions, taking advantage of their features and improvements. 4. :arrows_counterclockwise: Transition from Docker.io to Quay.io Container Registry: Due to Docker.io imposing lower download rate-limits, we are transitioning from Docker.io to Quay.io as our preferred container registry. 5. :bug: Bug Fixes and Stability Improvements: As part of our commitment to delivering a reliable and stable database solution, we have addressed reported issues and incorporated bug fixes in StackGres 1.5.0. The update offers enhanced stability, ensuring a smooth and uninterrupted experience for your critical workloads. ### Call to action We encourage all users to upgrade to StackGres 1.5.0 to take advantage of these exciting new features and enhancements. We value your feedback and suggestions, so please don't hesitate to reach out to our dedicated support team with any questions or comments. Join us on our [Slack](https://slack.stackgres.io/)/[Discord](https://discord.stackgres.io/) to share any comments or to get help from [StackGres Community](https://stackgres.io/community/). Thank you for choosing StackGres as your trusted Postgres database stack. We remain dedicated to providing top-notch performance, security, and scalability for your data needs. Thank you for all the issues created, ideas, and code contributions by the StackGres Community! Enjoy the power of StackGres 1.5.0! The StackGres Team ## :up: Upgrade To upgrade from a previous installation of the StackGres operator's helm chart you will have to upgrade the helm chart release. For more detailed information please refer to [our documentation](https://stackgres.io/doc/latest/install/helm/upgrade/#upgrade-operator). To upgrade StackGres operator's (upgrade only works starting from 1.1 version or above) Helm chart issue the following commands (replace namespace and release name if you used something different): `helm upgrade -n "stackgres" "stackgres-operator" https://stackgres.io/downloads/stackgres-k8s/stackgres/1.5.0/helm/stackgres-operator.tgz` #### :exclamation: IMPORTANT After each operator upgrade, remember to perform a security upgrade SGDbOps on each existing SGCluster after the operator has been upgraded. Example: ```yaml apiVersion: stackgres.io/v1 kind: SGDbOps metadata: name: secupgr spec: sgCluster: my-cluster op: securityUpgrade maxRetries: 1 securityUpgrade: method: InPlace ``` > :warning: NOTE: This release is incompatible with previous `alpha` or `beta` versions. Upgrading from those versions will require uninstalling completely StackGres including all clusters and StackGres CRDs (those in `stackgres.io` group) first. ### :construction: Known issues * Backups may be restored with inconsistencies when performed with a Postgres instance running on a different architecture ([#1539](https://gitlab.com/ongresinc/stackgres/-/issues/1539)) ## :clipboard: List of features and changes ### :sparkles: NEW FEATURES AND CHANGES * Added PostgreSQL 15.3, 15.2, 14.8, 14.7, 13.11, 13.10, 12.15, 12.14 * Added Babelfish for PostgreSQL 14.6, 13.9 * Added Patroni 3.0.2 * Support for Kubernetes 1.27 and 1.26 * Added SGShardedCluster CRD * Support to specify Patroni initial config * Support to set labels for Services (and Endpoints) * Support to set much more services parameters for Services * New sync-all and strict-sync-all to adjust the number of synchronous nodes to all replicas * Shift from docker.io to quay.io * Improved operator startup * Improved cluster startup * Enhanced cluster stats endpoint to fallback for cgroup v1 and v2 * Make CronJob generator for v1 or v1beta1 api versions dependent on Kubernetes version * Enable SSL by default on SGShardedCluster * Generate SSL certificate and private key if only enabled is specified * Change ssl configuration at run time * Use job backoffLimit to implements SGDbOps maxRetries * Improve current PgBouncer variable check query * Support for self signed certificate when using S3 compatible storage * Added priority class support * Changed Operator Bundle installation name to `stackgres-operator` * Cleanup CSR in operator helm chart * Return username key for cluster and sharded cluster info in REST API * Added sharded cluster stats REST API endpoints * Section sgshardedcluster.status.clusters returned by REST API the ordered names of the sgclusters that will be created * Updated base images for builds and release images #### Web Console * Support for Sharded Clusters * Support user-supplied sidecars for SGCluster services * Add loadBalancerIP to SGCluster and SGDistributedLogs * Support user-supplied sidecars for pods customVolumes, customInitContainers and customContainers * Enable SSL specs * Implemented cascading replication with WAL shipping * Implemented cascade replication from an external instance * Implemented cascade replication from a local sgcluster * Improve performance on REST API requests * Replace crd details table layouts with summary tree structure ### :bug: FIXES * ClusterBootstrapCompleted sent every few seconds * Empty backupPath for majorVersionUpgrade SGDbOps make the operation fail * Repeating restart ReducedImpact with 1 instance fails * Major version upgrade must never change extension version * Patroni do not clean up history after converting a cluster to standby cluster * Failed SGScript without retryOnError do not re-execute if version is changed * SGScript retry on error do not respect the backoff * Deleting a resource a SGDistributedLogs depends on should not be possible * pgbench SGDbOps fail if scale contains point * Helm Controller Manager selectors are too generic * Duplicate restartPolicy lines in the test operator template * REST API can-i do not return permissions for SGShardedCluster * REST API model is exposing parameters not present in the sharded cluster CRD * sgcluster REST API returns all pods for a sharded cluster * restart SGDbOps does not work on SGCluster when patroni scope is set * when SSL enable causes boostrap to fail and breaks synchronous replication * Sharded cluster primary service and shards service can not be set to NodePort or LoadBalancer * PgBouncer queries for postgres exporter are failing * Command `readlink /proc/$$/exe` may fail * During security upgrade cluster Pods are not found #### Web Console * Edit SGPostgresConfig form should only list custom parameters * Remove Advanced switch from Azure section * Summaries should not open unless all required fields are filled in * Backup scheduling inserting trailing zero in cron job minutes config * Script source not cleared when deleting a script * notValid class not being removed on Babelfish Experimental Feature * Unify click behavious when clicked on a switch and on its label * Cluster name on SGDbOp details is not clickable * Sidebar items hidden behind dialog popups * PITR date and time picker not working * Continue on SGScripts Error should not be visible if there are no Scripts set * Fix Extensions table layout * View Script button text and icon on Cluster Configuration tab have different behaviours * Hide empty sections on summaries * Disable Connection Pooling not working properly on Cluster Form * Object Storage selector on Cluster form shows all Object Storages from all Namespaces * Make sure crontab is shown on preferred timezone * Script content not shown on summary when set from a ConfigMap * Show secretKeySelectors for GCS service account json on edit mode * Review interceptors to REST API responses * SGBackups list won't load when start time is not present on a backup * Allow empty Backup path generate errors after update * Sets wrong path for SGCluster backup config * Edit SGPoolingConfig form should only list custom parameters * Replace instances dropdown with numeric input on SGCluster form * Service Account JSON not shown on Summary * Error when updating an SGCluster with a ConfigMap in managed SQL section * Cluster Summary opens and displays empty Custom Port properties * Only show form when ready to be edited ### :construction: KNOWN ISSUES * Backups may be restored with inconsistencies when performed with a Postgres instance running on a different architecture ([#1539](https://gitlab.com/ongresinc/stackgres/-/issues/1539)) ### :twisted_rightwards_arrows: [Full list of commits](https://gitlab.com/ongresinc/stackgres/-/commits/1.5.0)