Update jobservice base image (#136)

* use python 3.8 as base image for jobservic Signed-off-by: Khor Shu Heng <khor.heng@gojek.com> * fix linting Signed-off-by: Khor Shu Heng <khor.heng@gojek.com> * use official python image for test pipeline Signed-off-by: Khor Shu Heng <khor.heng@gojek.com> * Use python3 as default python version Signed-off-by: Khor Shu Heng <khor.heng@gojek.com> * simplify python package installation Signed-off-by: Khor Shu Heng <khor.heng@gojek.com> * Use JDK 11 for testing Signed-off-by: Khor Shu Heng <khor.heng@gojek.com> Co-authored-by: Khor Shu Heng <khor.heng@gojek.com>
feast-dev · Apr 12, 2022 · ae6309e · ae6309e
1 parent 6499fb2
commit ae6309e
Show file tree

Hide file tree

Showing 10 changed files with 12 additions and 13 deletions.
diff --git a/.github/workflows/pr.yml b/.github/workflows/pr.yml
@@ -52,7 +52,7 @@ jobs:
       PYSPARK_PYTHON: python3.7
     steps:
       - uses: actions/checkout@v2
-      - name: Set up JDK 8
+      - name: Set up JDK 11
         uses: actions/setup-java@v1
         with:
           java-version: '8'

diff --git a/Makefile b/Makefile
@@ -34,8 +34,7 @@ install-python-ci-dependencies:
 
 # Supports feast-dev repo master branch
 install-python: install-python-ci-dependencies
-	pip install --user --upgrade setuptools wheel
-	cd ${ROOT_DIR}/python; rm -rf dist; python setup.py install
+	pip install -e python
 
 lint-python:
 	cd ${ROOT_DIR}/python ; mypy feast_spark/ tests/

diff --git a/infra/docker/jobservice/Dockerfile b/infra/docker/jobservice/Dockerfile
@@ -1,4 +1,4 @@
-FROM jupyter/pyspark-notebook:399cbb986c6b
+FROM python:3.8
 
 USER root
 WORKDIR /app
@@ -13,9 +13,7 @@ RUN apt-get update && apt-get -y install make git wget
 # Install Feast SDK
 RUN git init .
 COPY README.md README.md
-RUN make install-python-ci-dependencies
 RUN make install-python
-RUN pip install "s3fs" "boto3" "urllib3>=1.25.4"
 
 #
 # Download grpc_health_probe to run health checks

diff --git a/infra/docker/spark/Dockerfile b/infra/docker/spark/Dockerfile
@@ -8,7 +8,7 @@ ARG VERSION=dev
 
 RUN REVISION=$VERSION make build-ingestion-jar-no-tests
 
-FROM gcr.io/kf-feast/spark-py:3.0.2 as runtime
+FROM gcr.io/kf-feast/feast-spark-base:v3.0.3 as runtime
 
 ARG VERSION=dev
 

diff --git a/infra/docker/spark/dev.Dockerfile b/infra/docker/spark/dev.Dockerfile
@@ -1,4 +1,4 @@
-FROM gcr.io/kf-feast/spark-py:3.0.2 as runtime
+FROM gcr.io/kf-feast/feast-spark-base:v3.0.3 as runtime
 
 ARG VERSION=dev
 

diff --git a/infra/scripts/test-end-to-end-sparkop.sh b/infra/scripts/test-end-to-end-sparkop.sh
@@ -63,7 +63,7 @@ kubectl delete pod -n "$NAMESPACE" ci-test-runner 2>/dev/null || true
 kubectl run -n "$NAMESPACE" -i ci-test-runner  \
     --pod-running-timeout=5m \
     --restart=Never \
-    --image="${DOCKER_REPOSITORY}/feast-ci:latest" \
+    --image="python:3.7" \
     --env="FEAST_TELEMETRY=false" \
     --env="DISABLE_FEAST_SERVICE_FIXTURES=1" \
     --env="DISABLE_SERVICE_FIXTURES=1" \

diff --git a/python/feast_spark/pyspark/historical_feature_retrieval_job.py b/python/feast_spark/pyspark/historical_feature_retrieval_job.py
@@ -331,7 +331,7 @@ def _map_column(df: DataFrame, col_mapping: Dict[str, str]):
         else:
             projection[col_name] = expr(col_mapping.get(col_name, col_name))
 
-    return df.select([c.alias(a) for a, c in projection.items()])
+    return df.select([c.alias(str(a)) for a, c in projection.items()])
 
 
 def as_of_join(
@@ -545,7 +545,7 @@ def join_entity_to_feature_tables(
         joined_df = as_of_join(
             joined_df, entity_event_timestamp_column, feature_table_df, feature_table,
         )
-        if SparkContext._active_spark_context._jsc.sc().getCheckpointDir().nonEmpty():
+        if SparkContext._active_spark_context._jsc.sc().getCheckpointDir().nonEmpty():  # type: ignore
             joined_df = joined_df.checkpoint()
 
     return joined_df
@@ -614,7 +614,7 @@ def filter_feature_table_by_time_range(
         )
         .where(col("distance") == col("min_distance"))
     )
-    if SparkContext._active_spark_context._jsc.sc().getCheckpointDir().nonEmpty():
+    if SparkContext._active_spark_context._jsc.sc().getCheckpointDir().nonEmpty():  # type: ignore
         time_range_filtered_df = time_range_filtered_df.checkpoint()
 
     return time_range_filtered_df

diff --git a/python/requirements-ci.txt b/python/requirements-ci.txt
@@ -24,3 +24,4 @@ great-expectations==0.13.2
 adlfs==0.5.9
 redis==4.1.*
 Jinja2==3.0.3
+croniter==1.*
diff --git a/python/setup.py b/python/setup.py
@@ -53,6 +53,7 @@
     "mypy-protobuf==2.5",
     "croniter==1.*",
     "redis==4.1.*",
+    "prometheus-client==0.14.0"
 ]
 
 # README file from Feast repo root directory

diff --git a/spark/ingestion/src/test/resources/python/setup.sh b/spark/ingestion/src/test/resources/python/setup.sh
@@ -14,5 +14,5 @@ fi
 
 # 2. Pickle python udf
 cd $CURRENT_PATH
-pip3 install great-expectations==0.13.2 setuptools pyspark==3.0.1
+pip3 install great-expectations==0.13.2 setuptools pyspark==3.0.1 Jinja2==3.0.3 pyarrow==2.0.0
 python3 udf.py $DESTINATION/udf.pickle