Upgrade arrow-cpp and pyarrow to 7.0.0

[galipremsagar]

Changes needed for: NVIDIA/cudf#10503

[galipremsagar]

rerun tests

[galipremsagar]

rerun tests

[ajschmidt8]

rerun tests

[ajschmidt8]

The PR build job is running below for dfaf1d6 3463681. I'm not sure why the status is not being reported on this PR. Could be related to the GitHub issues that have been occurring over the last few days

• https://gpuci.gpuopenanalytics.com/job/rapidsai/job/integration/job/prb/job/integration-prb/1006/

[ajschmidt8]

Actually, on second thought, after checking the logs. It seems that our comments triggered CI for commit 3463681 instead of the latest commit 🤔. Not sure what's going on here. Likely still GitHub related though

[galipremsagar]

It also appears that 22.06 package for xgboost is not found: https://gpuci.gpuopenanalytics.com/job/rapidsai/job/integration/job/prb/job/meta-pkg-nightly-build-arm64/CONDA_CONFIG_FILE=conda%2Frecipes%2Fversions.yaml,CONDA_USERNAME=rapidsai-nightly,CUDA_VER=11.5,PYTHON_VER=3.8,RAPIDS_VER=22.06.00a/215/console

[ajschmidt8]

Actually, on second thought, after checking the logs. It seems that our comments triggered CI for commit 3463681 instead of the latest commit thinking. Not sure what's going on here. Likely still GitHub related though

I'll work on getting that published.

[ajschmidt8]

@galipremsagar, it seems that 22.06 packages are now available: https://anaconda.org/rapidsai-nightly/xgboost/files

[galipremsagar]

@galipremsagar, it seems that 22.06 packages are now available: https://anaconda.org/rapidsai-nightly/xgboost/files

Thanks for working on bringing them up AJ!

[ajschmidt8]

@galipremsagar. correction! still waiting on ARM packages

[galipremsagar]

rerun tests

[galipremsagar]

Looks like 22.06 packages aren't published for cugraph & cuml:

14:27:32 conda.exceptions.ResolvePackageNotFound: 
14:27:32   - cuml=22.06
14:27:32   - cugraph=22.06
14:27:32   - pylibcugraph=22.06
14:27:32   - libcugraph_etl=22.06

[galipremsagar]

rerun tests

[ajschmidt8]

admin merging. the relevant tests have passed and NVIDIA/cudf#10503 is passing CI successfully with the new version.