extras-requirements.txt

# In our docker images we're copying the requirements files and pip installing them before copying and installing the
# whole code. We're doing that to leverage docker build cache so that we won't need to actually install all the
# requirements for every change in the code (which causes a cache miss in the copy command for the whole code) but we'll
# get "requirement is already satisfied" which causes the pip install of the mlrun package itself to be very quick.
# The extras requirements definition sits in the setup.py itself, therefore without this file these requirements were
# installed only in the last phase of installing mlrun, which causes this step to be longer and lowers the effectiveness
# of the above trick. To overcome this we have this file, which is a copy of all the requirements defined in the extras
# in setup.py so that we'll be able to copy and install this in the layer with all other requirements making the last
# layer (which is most commonly being re-built) as thin as possible
# we have a test test_extras_requirement_file_aligned to verify this file is aligned to setup.py
# The versions of gcsfs and s3fs should be identical to the fsspec version;
# Therefore, they must be equal to each other.
s3fs==2023.9.2
gcsfs==2023.9.2
boto3>=1.28.0,<1.29.0
aiobotocore>=2.5.0,<2.8
# https://github.com/Azure/azure-sdk-for-python/issues/24765#issuecomment-1150310498
msrest~=0.6.21
azure-core~=1.24
# This version is required in order to support the self._strip_protocol
# function in Azure Blob File System functions (such as _ls and more).
adlfs==2023.9.0
azure-identity~=1.5
azure-keyvault-secrets~=4.2
# cryptography>=39, which is required by azure, needs this, or else we get
# AttributeError: module 'lib' has no attribute 'OpenSSL_add_all_algorithms' (ML-3471)
pyopenssl>=23
bokeh~=2.4, >=2.4.2
# plotly artifact body in 5.12.0 may contain chars that are not encodable in 'latin-1' encoding
# so, it cannot be logged as artifact (raised UnicodeEncode error - ML-3255)
plotly~=5.4, <5.12.0
# To support python 11, google-cloud-bigquery requires grpcio>=1.49.1, which is incompatible with the grpcio version
# required by frames (because it upgrades protobuf from 3.x to 4.x, breaking binary compatibility)
google-cloud-bigquery[pandas, bqstorage]==3.14.1
kafka-python~=2.0
avro~=1.11
redis~=4.3
graphviz~=0.20.0
mlflow~=2.8
# any newer version then 0.13 should be tested with adlfs,gcsfs and s3fs installation.
# we experience a different in google-auth package in databricks-sdk>0.13
databricks-sdk~=0.13.0
# sqlalchemy version should be aligned with mlrun-api due to common use with the model monitoring database
sqlalchemy~=1.4
dask~=2023.9.0
distributed~=2023.9.0