Releases: sematic-ai/sematic
Releases Β· sematic-ai/sematic
0.16.0
What's Changed
- Support enums in Sematic function annotations
- Allow specifying Kubernetes tolerations for cloud jobs
- Redesign log view UI
Full Changelog: v0.15.1...v0.16.0
0.15.1
What's Changed
- Ensure log ingestion happens when using bazel cloud jobs and
dev=False
- Avoid spamming the server with requests for logs for incomplete runs
without logs
Full Changelog: v0.15.0...v0.15.1
0.15.0
What's Changed
- Display logs for cloud jobs in the UI. Note: This feature will require your server to have the ability to list objects in the S3 bucket you're using for Sematic storage.
- Ensure the push of docker images by bazel works with docker credential helpers
Full Changelog: v0.14.0...v0.15.0
0.14.0
What's Changed
In prior versions, anyone who needed to submit a pipeline execution had to get set up to access the Kubernetes cluster where the jobs would run. As of this release, users only need to get set up to access a Sematic server, and the server will handle all communication with Kubernetes. This improves user experience as well as security. It does mean that anyone running cloud execution may need to change their deployment so that the server can access Kubernetes. We've updated the deployment docs with instructions and created a Helm chart to ease deployment. As a summary of this release:
- BREAKING CHANGE: For cloud execution, users submitting pipelines for execution no longer need Kubernetes access. Users who have deployed the Sematic server using docker will need to either (a) switch to deploying the server on Kubernetes or (b) use their Sematic deployment only for metadata tracking instead of metadata tracking + cloud execution.
- Remove job environment variables from resolution API response
- Lower the probability of evictions of resolution jobs and improve handling of that situation
Full Changelog: v0.13.0...v0.14.0
0.13.0
What's Changed
- Bugfix for dicitonary visualization
- When running in cloud mode, have the server log to stdout
- Enable mounting Kubernetes secrets into the containers for sematic funcs
Full Changelog: v0.12.0...v0.13.0
0.12.0
What's Changed
- BREAKING CHANGE: Allow specifying resource requirements for Kubernetes jobs.
KubernetesResourceRequirements
has a new required field,requests
- Add
has_container_image()
API for better control over launch workflows - Fix a bug with direct run urls
Full Changelog: v0.11.0...v0.12.0
0.11.0
What's Changed
- Add a
get_artifact_value
API to retreive artifacts by ID - Add ability to link to individual runs on the UI, not just pipelines
- Add option to serve local server from 0.0.0.0
- Add capability to use different API URLs for local vs remote client usage
- Enable environment variable to change Sematic config directory
- Add exceptions and other improved app logs to the server log files
- Solve a bug with displaying empty lists in the UI
Full Changelog: v0.10.0...v0.11.0
0.10.0
What's Changed
- Added support for python 3.8
- Friendly error message for unsupported Python version
- Friendly error message when clients don't match server version
Full Changelog: v0.9.0...v0.10.0
0.9.0
What's Changed
- Baseline Grafana integration for log panels by @neutralino1 in #72
Full Changelog: v0.8.0...v0.9.0
0.8.0
What's Changed
- Persist stack traces and display in UI by @neutralino1 in #71
Full Changelog: v0.7.0...v0.8.0