V2.0

TensorFlow deployment for local Kubernetes cluster.

A TensorFlow parameter server deployment for Kubernetes has been created and the corresponding base script also has been added.

What's Changed

• The tran_tf_ps.py script has been added for table-like data and for training with images.
• The architecture for health example data and SFS sensor regression model has been created.
• MySQL deployment for GCP is working.
• Spark for GCP deployment is working.
• Spark connection for GCP is working.
• Parameter server is working for local servers (GCP has not been tested). The architecture is the same as the Spark and MySQL deployments, so, must be working (is ready for tests).

Full Changelog: V1.1...V2.0

Private Spark Kubernetes Cluster

The Private Kubernetes Cluster for Spark Workloads is done.

What's Changed

• Add internal load balancer annotation to Spark master service. by @greg-ogs in #9

Private Kubernetes Cluster

• The Kubernetes cluster architecture was successfully crated.
• The load balancer service is configurated to use private VPC ip addresses only.
• The access is only accessible trough bastion vm.

Spark Workloads

The spark deployments allow the pyspark workload in this cluster.

Full Changelog: V1.0...V1.1

V1.0

What's Changed

*Infrastructure are tested and is working.
*Python spark example is working.

• Testing workload using Python by @alex-ogs in #7

New Contributors

• @alex-ogs made their first contribution in #7

Full Changelog: Alpha-v0.1...V1.0

Alpha-v0.1

Kubernetes cluster for spark workload ready for tests with synthetic data and knew datasets.

What's Changed

• Add CodeQL workflow for enhanced code analysis by @greg-ogs in #1
• Testings for Kubernetes cluster in GKE ready by @greg-ogs in #6

New Contributors

• @alex-ogs made their first contribution.

Full Changelog: https://github.com/greg-ogs/PySpark-TF-GKE/commits/Alpha-v0.1