Skip to content

Releases: opendatacube/datacube-ows

OWS 1.8.3 release March 2021

11 Mar 23:18
fb2cf24
Compare
Choose a tag to compare
  • Generalised handling of WMTS tile matrix sets (#452)
  • Progressive cache control headers (#476)
  • Support for multi-product masking flags. (#499)
  • Greatly improved test coverage (various)
  • Many bug-fixes, documentation updates and minor enhancements (various)

Config enhancement and bug fixes

26 Oct 04:35
4cb7493
Compare
Choose a tag to compare
  • Config inheritance for layers and styles.
  • CRS aliases
  • Enhanced band util functions.
  • Query stats parameter.
  • Stand-alone config parsing/validating tool.
  • Cleaner internal APIs, improved test coverage, and bug fixes.

Bug fixes and performance enhancements

18 Aug 05:16
87de187
Compare
Choose a tag to compare
  1. Bug fixes
  2. Performance enhancements - most notable using materialised views for spatio-temporal DB searches.
  3. Improved testing and documentation.

1.8.0 release

10 Jun 03:45
Compare
Choose a tag to compare
  • now bases off odc 1.8.x
  • supports for wcs2.0
  • update_ranges changes (see below)

Update_ranges.py now uses the spatio-temporal materialised views proposed for odc 2.0.

You will need to run update_ranges.py --schema --role myrole once after upgrading to create the materialised views.

Thereafter the materialised views can be periodically refreshed with update_ranges.py --views.

The --product, --all and --multiproduct options are now deprecated - you simply need to list the OWS layers that you wish to update. (ODC product names can also be used).

The default refresh mode is a non-blocking/concurrent refresh. - i.e. the refresh is calculated in the background and takes effect when complete, with the pre-refresh data readable in the mean time. Therefore to update the OWS ranges for layer "mylayer" with a guarantee to have the latest available data it would be necessary to run:

update_ranges.py --views --blocking
update_ranges.py mylayer

Running --views without the --blocking flag would mean that the update of mylayer could use old data as the refresh could still be running in the background.

Fix for https://github.com/opendatacube/datacube-ows/issues/258

01 May 03:47
c9392e2
Compare
Choose a tag to compare
Merge pull request #259 from opendatacube/release-action

Try github.event_name

Release latest patch

30 Apr 23:44
eee9158
Compare
Choose a tag to compare

0.14.3 failed to create a docker image, re-release the same version
Rerun after change to github.event_name

Patch

30 Apr 02:29
ade0bfe
Compare
Choose a tag to compare

Fixes a few issues from recent data change

Conformance wcs fixes

29 Apr 04:41
bbef3f7
Compare
Choose a tag to compare
Merge pull request #253 from opendatacube/fix-pylint

Fix pylint version to 2.4.4

Release using github actions

20 Apr 06:02
Compare
Choose a tag to compare
0.14.1

run actions on release

Docker image rebuild

20 Apr 05:54
d33f0c2
Compare
Choose a tag to compare

In this release we re-worked the docker image to be based on opendatacube/geobase, and to use multi-stage builds to create a slimmer docker image. The images should be compatible with previous builds, but there are some key differences:

  • no index / setup scripts present in the image
  • support datacube-ows-update instead of running the python script directly
  • slimmer image (less Linux utils for increased security)
  • run as non-root
  • no startup script (improves compatibility with Kubernetes, by passing SIGTERM to gunicorn when pod is asked to shutdown)