Closed mrocklin closed 1 year ago
+1 -- thanks for handling @mrocklin. Release procedures are here https://github.com/dask/dask/blob/main/docs/release-procedure.md. The only thing that's changed since the last time you've released is you need to be sure to update the dask
/ distributed
version pinning since they are now tightly pinned (i.e. make sure dask=2022.10.2
pins distributed=2022.10.2
and vice versa). Here's an example of that happening in the release last Friday https://github.com/dask/dask/commit/67beb3ec71b55f3fc2f82b3b7aa6be20b06085a5 and https://github.com/dask/distributed/commit/1db2595266fadd9b9163f6802707bdcc53c71214
Cool. Thanks for such quick response :)
Proceding with the release now
Pushed to PyPI
Build is running for Airflow : https://github.com/apache/airflow/pull/27415
And Mergeed! All tests with dask passed. Thanks for super quick reaction @mrocklin and the Dask team.
No worries. My apologies for the confusion.
On Mon, Oct 31, 2022 at 12:31 PM Jarek Potiuk @.***> wrote:
And Mergeed! All tests with dask passed. Thanks for super quick reaction @mrocklin https://github.com/mrocklin and the Dask team.
— Reply to this email directly, view it on GitHub https://github.com/dask/community/issues/284#issuecomment-1297432297, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACKZTC3FGMBUXQRRGEGXVLWF766XANCNFSM6AAAAAARTGYJR4 . You are receiving this because you were mentioned.Message ID: @.***>
Looks like conda-forge didn't automatically open PRs to bump to 2022.10.2, so opened them manually:
cc @jakirkham in case you have any idea why this might be? My first guess was that the dependency metadata didn't change at all between releases
The bot that opens these PRs does so on a cron job. So it depends a bit on when the release happened and when the last cron job was run.
Anyways all of those packages are published as are the new Docker images ( https://github.com/dask/dask-docker/pull/258 ). Closing as completed. Though if I've missed something, please feel free to reopen.
Thanks all! 🙏
Thanks all!
Everyone without Bokeh is currently getting failing clusters, as reported at https://github.com/dask/distributed/issues/7227 and https://github.com/dask/dask/issues/9609 and fixed in https://github.com/dask/distributed/pull/7230 by @oliverholworthy
I'm going to be pushing out a hotfix now.