ray-packages v2.1.0

[regro-cf-autotick-bot]

It is very likely that the current package version for this feedstock is out of date.

Checklist before merging this PR:

• Dependencies have been updated if changed: see upstream
• Tests have passed
• Updated license if changed and license_file is packaged

Information about this PR:

1. Feel free to push to the bot's branch to update this PR if needed.
2. The bot will almost always only open one PR per version.
3. The bot will stop issuing PRs if more than 3 version bump PRs generated by the bot are open. If you don't want to package a particular version please close the PR.
4. If you want these PRs to be merged automatically, make an issue with @conda-forge-admin,please add bot automerge in the title and merge the resulting PR. This command will add our bot automerge feature to your feedstock.
5. If this PR was opened in error or needs to be updated please add the bot-rerun label to this PR. The bot will close this PR and schedule another one. If you do not have permissions to add this label, you can use the phrase @conda-forge-admin, please rerun bot in a PR comment to have the conda-forge-admin add it for you.

Closes: #76

Pending Dependency Version Updates

Here is a list of all the pending dependency version updates for this repo. Please double check all dependencies before merging.

Name	Upstream Version	Current Version
openjdk	19.0.1
ray-packages	2.1.0

Dependency Analysis

We couldn't run dependency analysis due to an internal error in the bot. :/ Help is very welcome!

_{This PR was created by the regro-cf-autotick-bot. The regro-cf-autotick-bot is a service to automatically track the dependency graph, migrate packages, and propose package version updates for conda-forge. Feel free to drop us a line if there are any issues! This PR was generated by https://github.com/regro/autotick-bot/actions/runs/3416647520, please use this URL for debugging.}

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

I do have some suggestions for making it better though...

For recipe:

• Recipes should usually depend on matplotlib-base as opposed to matplotlib so that runtime environments do not require large packages like qt.

[h-vetinari]

This will have to wait for #77. Once ready, we also need to unpin grpc completely. The previous problems were solved by ray-project/ray#28354, though upstream stayed conservative with a <=1.48.1 pin. However, conda-forge is already on 1.49 (and 1.50 is out already), so to avoid resolvability issues (and in the absence of strong reasons like there were previously), we should unpin grpc, at the very least to include 1.49.

Edit: NVM, the cap was explicitly lifted in ray-project/ray@9505cd1

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[h-vetinari]

@conda-forge-admin, please rerender

[github-actions]

Hi! This is the friendly automated conda-forge-webservice.

I tried to rerender for you, but it looks like there was nothing to do.

This message was generated by GitHub actions workflow run https://github.com/conda-forge/ray-packages-feedstock/actions/runs/3418176799.

[h-vetinari]

@mattip @vnlitvinov
FYI, The redis-patch needs updating yet again... 😑

[h-vetinari]

@vnlitvinov
I reworked the redis-patch to build on top of the existing one in the ray sources. I think that makes for a much more readable patch, see e.g. here. I've also taken your suggestion to update ray-patching.md.

PTAL

PS. I'd like to fix the OSX (see #80) before we merge this.

[mattip]

I don't think the test failure is related to inability to start workers. Further up there is

ERROR services.py:1403 -- Failed to start the dashboard: \
    Failed to start the dashboard, return code -11

Maybe there is something off elsewhere?

[mattip]

In order to separate dashboard failures from other issues, could you add a test

- python -c "import ray; ray.init(include_dashboard=False)"

before this line

ray-packages-feedstock/recipe/meta.yaml

Line 97 in 7002f2a

- python -c "import ray; ray.init()"

[h-vetinari]

I tried to follow up an unlikely gut-feeling, and it seems indeed that a patch-version of protobuf might have broken things. In particular, things break with 4.21.9 but run through with 4.21.6 (in both cases with grpc 1.49.1)

[h-vetinari]

I tried to follow up an unlikely gut-feeling, and it seems indeed that a patch-version of protobuf might have broken things. In particular, things break with 4.21.9 but run through with 4.21.6 (in both cases with grpc 1.49.1)

Nevermind, pinning protobuf seems to have backed the resolver into some weird corners. Let's try again even more explicitly.

[vnlitvinov]

xref modin-project/modin#5208 also with Ray 2.1.0 (from pip, but other deps from conda-forge).
Seems like some dependencies are bad.

[h-vetinari]

OK, I wasn't thinking that the host-dep-at-build-time gets picked up as a lower bound, so grpcio 1.49.1 + protobuf 4.21.6 is impossible to co-install (grpc was built against libprotobuf 3.21.9, meaning it needs >=3.21.9 at runtime; this corresponds to protobuf 4.21.9 after protobuf's recent versioning scheme change)

[mattip]

What would you think about extending the summary to mention all the different packages provided to make the summary at the various links like https://anaconda.org/conda-forge/ray-default more informative? I would like to link to it from https://docs.ray.io/en/latest/ray-overview/installation.html#installing-ray-with-anaconda as part of reworking that documentation stanza.

    - licenses/redis-hiredis-COPYING.txt
     - licenses/tporadowski-redis-license.txt
     - licenses/zlib-LICENSE.txt
-  summary: Ray is a fast and simple framework for building and running distributed applications.
+  summary: |
+    Ray is a fast and simple framework for building and running distributed
+    applications. It is split into ray-core, ray-default, ray-dashboard,
+    ray-serve, ray-rllib, ray-k8s, ray-data, ray-tune, and ray-all packages.
   description: |
     Ray is a fast and simple framework for building and running
     distributed applications.

[h-vetinari]

What would you think about extending the summary

Fine by me, though personally I'd extend the description rather than the summary (no strong preference however).

[h-vetinari]

Do we want to publish this with a pin for grpcio<1.49 for the moment?

[mattip]

A pin sounds like the most practical solution indeed.

[mattip]

CI is passing. I will merge tomorrow if there are no objections.

[vnlitvinov]

Looks good overall, though I would like to mimic dependencies from setup.py in a more close manner.

[vnlitvinov]

Minor suggestion left, can go without if you wish. Thanks!

[vnlitvinov]

LGTM!

[github-actions]

Hi! This is the friendly conda-forge automerge bot!

I considered the following status checks when analyzing this PR:

• linter: passed
• azure: passed

Thus the PR was passing and merged! Have a great day!