Releases: faradayio/falconeri
Releases · faradayio/falconeri
v0.2.13: Miscellaneous polishing
Added
- Wrote some basic developer documentation to supplement the
justfile
s. - Allow specifying
--falconerid-log-level
forfalconeri deploy
. This uses standardRUST_LOG
syntax, as described in the CLI help.
Fixed
- Cleaned up tracing output a bit.
- Switched to using
rustls
for HTTPS. Database connections still indirectly require OpenSSL thanks tolibpq
.
v0.2.12: Fix binary builds
Fixed
- Attempt to fix TravisCI binary releases.
v0.2.11: Update dependencies and add timeouts
Added
- Don't show interactive progress bar when uploading outputs.
- Support
job_timeout
in pipeline schemas. This allows you to specify when an entire job should be stopped, even if it isn't done. Values include "300s", "2h", "2d", etc. - Add much better tracing support when
RUST_LOG=trace
is passed.
Changed
- We update most of our dependencies, including Rust libraries and our Docker base images. But this shouldn't affect normal use.
Fixed
- Set
ttlSecondsAfterFinished
to 1 day so that old jobs don't hang around forever on the backplane wasting storage.
v0.2.10: Auto-run migrations on `falconerid` startup
This makes it easier to deploy `falconerid`, though it does prevent us from writing complex migrations that affect a running cluster.
v0.2.9: Fix `falconeri job retry`
This subcommand was broken in 0.2.8.
v0.2.8: Improve `falconeri deploy` for production APIs
This release contains a number of fixes to the deployment manifest generated by `falconeri deploy`, as well as a new `--skip-secret` option that will prevent the secret from being output. The documentation on EKS `Ingress` resources has also been updated.
v0.2.7: Add REST API support
`falconerid` now supports a limited, not-yet-documented REST API which can be used to run jobs and retry them when they fail.
v0.2.6: Official 0.2.x release with everything
Big changes since 0.1.x: - HTTP-based communication between workers and falconerid, improving scalability. - `FALCONERI_PROXY_HOST` environment variable to make it easier to run `falconeri` in production environments. If you're running 0.1.x, you will need to run `falconeri undeploy` with the old version, and `falconeri deploy` with the new.
v0.1.9: Add more input and glob types
We also updated to Rust 2018.
v0.1.8: Retry DB connection failures
This may help when running at scale on the cloud.