Releases: vanvalenlab/kiosk-console
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
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 theautoscaler
,redis-janitor
, andfrontend
. - Custom Prometheus metrics have been added to the
prometheus-operator
(polaris_consumer_key_ratio
andpolaris_zip_consumer_key_ratio
) - This PR adds the new consumer along with its complementary zip-consumer
- Bumps
kiosk-redis-consumer
to 0.14.0 andkiosk-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.
1.6.0
1.5.1
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 aninitContainer
. (#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 RedisinitContainer
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
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
withgomplate
and unnecessary environment variables (#293). - Cleaned up and updated prometheus helmfiles and improved the custom prometheus rules using labels. (#304)
- Refactor
kiosk
tokiosk-console
. (#295, #302) - Updated
autoscaler
to0.4.1
. (#308) - Updated
frontend
to0.4.1
. (#310) - Updated
redis
to10.5.7
. (#307, #315) - Updated
openvpn
to4.2.1
. (#301) - Updated
redis-janitor
to0.3.2
(#339) - Updated
redis-consumer
to0.6.0
. (#337) - Updated
benchmarking
to0.3.0
(#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
to1.14
.