1.8.1

chore: upgrade geodesic base image version (#464) @willgraf (#465)

• chore: update geodesic to latest version.

Update some env vars and scripts too.

• fix: yq has new syntax

• feat: use new gke auth plugin to silence warnings.

• fix: update HPA API definition

• fix: drop MAKE_INCLUDES changes, confusing make

• chore: silence user output for IAM commands.

• fix: update HPA metrics for v2beta2 API.

• fix: only install plugin if it is enabled.

Disabled by default as the plugin is recommended but not required.

1.8.0

🚀 Features

Adding Polaris consumer and Polaris zip consumer #448

Polaris super consumer #450

chore: update to Kubernetes 1.22 #461

1.7.0

🚀 Features

Polaris super consumer @elaubsch (#450)

This PR creates the Polaris super consumer. The Polaris queue and consumer previously handled spot detection tasks but have been refactored to the spots queue and consumer. The changes to maximum batch size are an attempt to address #449, but the issue is persisting. A temporary solution was implemented in vanvalenlab/kiosk-redis-consumer#183.

Adding Polaris consumer and Polaris zip consumer @elaubsch (#448)

The Polaris consumer performs spot detection (see deepcell_spots.applications.SpotDetection). There are several files that are changed to add the new job:

• prometheus-redis-exporter adds the job queue to be exported to Prometheus.
• The QUEUES environment variable has been updated in the autoscaler, redis-janitor, and frontend.
• Custom Prometheus metrics have been added to the prometheus-operator (polaris_consumer_key_ratio and polaris_zip_consumer_key_ratio)
• This PR adds the new consumer along with its complementary zip-consumer
• Bumps kiosk-redis-consumer to 0.14.0 and kiosk-frontend to 0.9.0

🧰 Maintenance

Update segmentation and tracking models for redeployment @msschwartz21 (#453)

This PR updates the segmentation and tracking models for redeployment. The updated frontend image supports viewers for both mesmer and polaris. The updated consumer image supports multi-batch segmentation predictions.

Update master with Polaris consumer updates made to stable @msschwartz21 (#451)

This PR should bring master up to date with the changes made to stable (#448, #450). After this PR is complete, stable will be deleted so that we only need to maintain the master branch.

Add release drafter workflow @msschwartz21 (#452)

This sets up the automated release drafter workflow to simplify maintenance for this repo.

📚️ Documentation

Bump version to 1.7.0 in docs @msschwartz21 (#454)

In anticipation of a new release of the kiosk that includes the updates to support polaris.

Update copyright year and documented version numbers. @willgraf (#446)

1.6.0

What's Changed

• Update KUBERNETES_VERSION to 1.21 by @willgraf in #442
• Update consumers to 0.13.0 and refactor "tracking" to "caliban" by @willgraf in #442
• Update tf-serving to v0.5.0, using TensorFlow Serving v2.5.3 by @willgraf in #442
• Bump frontend version to 0.8.2 by @willgraf in #444

1.5.1

1.5.1

Bugfixes

• Update KUBERNETES_VERSION to 1.17.x (Fixes #428)
• Bump autoscaler to 0.6.0
• Bump redis-janitor to 0.5.0

1.5.0

1.5.0

Features

• Add new consumer and zip consumer for "multiplex" jobs. (#420)
• Bump tracking model to TensorFlow 2 based version. (#426)

Bugfixes

• Fixed bug in unescaped annotations in helm templates (#421)

Breaking Changes

• Upgrade default GKE version to 1.16 (#412)
• Update autoscaler and janitor to the latest release for better compatibility with 1.16. (#419)
• Update bucket-monitor helm chart with unified STORAGE_BUCKET environment variable. (#414)
• Update helm charts to use single STORAGE_BUCKET environment variable including bucket protocol. (#414, #417)
• Refactor multiplex consumers to Mesmer consumers (#422)

1.4.0

1.4.0

Features

• Set up an AlertManager with slack receiver support. (#317)
• Support TLS traffic with cert-manager. (#357)
• Create tf-serving configuration files using an initContainer. (#382)
• Template frontend ingress annotations using CERTIFICATE_MANAGER_ENABLED (#383)
• Add a screenshot of a successfully created cluster to docs. (#388)
• Migrate CI/CD from TravisCI to GitHub Actions (#394)

Bugfixes

• Fix frontend ingress issue when no hosts are provided. (#381)
• Update the make helmfile/destroy/all command for helm 3. (#389)
• Install procps in Redis initContainer to give access to sysctl. (#390)
• Skip gke/destroy/node-pools task during cluster teardown. (#407)
• Migrate stable helm chart repo to archived URL. (#399)

Breaking Changes

• Change the Redis helm chart repo to bitnami (#393)
• Remove deprecated and unused charts (#398)
• Move HPA definitions to inside helm charts (#395, #396, #397)
• Move the prometheus-redis-exporter script to a chart using incubator/raw. (#405)
• Update helmfile defaults for faster helm deployments. (#404)
• Use kubectl del pvc instead of deleting all pds with the cluster name. (#406)
• Bump openvpn to 4.2.3 (#385)
• Bump redis-consumer version to 0.8.3 (#409)
• Upgrade tf-serving chart to 0.3.0 for application version 0.4.0 (#392)

1.3.0

1.3.0

Features

• Upgrade helm2 to helm3 for better support with newer Kubernetes versions.
• Fixed the README to include the CLI and the ImageJ Plugin.

v1.2.1

v1.2.1

Bugfixes

• Fixed documentation badge (#354)
• Bump redis-consumer versions to 0.6.1 (#355).
• Select us-west1 as default project region if none is set (#358).
• Clarify quota documentation (#359, #362).
• Fix stale project names (#364).
• Destroy prometheus-operator CRDs when helm chart is deleted (#365).
• Warn the user about multizone egress fees (#366).
• Prune deprecated helmfiles (#368).
• Add documentation on cluster benchmarking (#374)
• Update default KUBERNETES_VERSION to 1.15 (#376)

v1.2.0

v1.2.0

This release updates several dependencies and improves the custom Prometheus rules with better label matching. The repository has been renamed kiosk-console from kiosk to better illustrate this project as a part of the whole.

Features

• Template conf/patches/hpa.yaml with gomplate and unnecessary environment variables (#293).
• Cleaned up and updated prometheus helmfiles and improved the custom prometheus rules using labels. (#304)
• Refactor kiosk to kiosk-console. (#295, #302)
• Updated autoscaler to 0.4.1. (#308)
• Updated frontend to 0.4.1. (#310)
• Updated redis to 10.5.7. (#307, #315)
• Updated openvpn to 4.2.1. (#301)
• Updated redis-janitor to 0.3.2 (#339)
• Updated redis-consumer to 0.6.0. (#337)
• Updated benchmarking to 0.3.0 (#341)
• Updated README with a logo and banner (#332)
• Added new 4GPU menu option (#333)
• Added unit tests and make all tests use the Docker image. (#320, #326)
• Warn the user if the bucket is not in the same region as the cluster (#329)
• Changed zip consumer HPA target to 2 (#334)
• Enabled single-zone clusters from the menu (#340)
• Pin KUBERNETES_VERSION to 1.14.

Bugfixes

• Updated default account firewall setting to 100. (#297)
• patches/gke-helmfile-deployment.sh now uses yq and helmfile build to dynamically deploy all releases in a given helmfile. (#300)
• Fixed stale environment variable references in addons/cluster-autoscaler.yaml. (#305)