v3.3.1

v3.3.1 is a routine refresh with no new features.

Improvements

• None

Changes

• Updated version tag in container Docker files
• Updated version tag in docker-compose.yml for dbmanager and collector
• Updated containers on Docker Hub

Fixes

• Upstream: Python requests library version updated to 2.32.0 (fixes minor security bug in requests)

v3.2.0

v3.2.0 makes it easier to build, use and maintain E-Series-related (collector & dbmanager) containers and leaves the rest to the user in line with the ideas behind this fork.

Release v3.2.0 has changes that make it easier to deploy EPA on Kubernetes, but it does not contain new features for Docker Compose users, so Docker-Compose users can continue using v3.1.0 (especially if they encounter problems deploying v3.2.0)

Changes and improvements

Improvements

• New ./kubernetes subdirectory with detailed instructions on how to deploy dbmanager and collector on Kubernetes
• Sample configuration for InfluxDB v1 and Grafana v8 in ./kubernetes/samples

Changes

• Small changes to dbmanager
• Many changes to existing README files
• New diagram of EPA components
• For convenience, YAML files use Docker Hub containers

Fixes

• None

v3.1.0

v3.1.0 makes it easier to build, use and maintain E-Series-related (collector) containers and leave the rest to the user, in line with the ideas behind this fork.

Changes and improvements

Improvements

• Change ./epa/Makefile and other files in that directory so that collector and WSP containers are no longer built, making InfluxDB and Grafana container builds faster.
• New ./collector/Makefile, simple Makefile for Collector and dbmanager.

Changes

• Remove all WSP-related functions out of collector.py
• ./collector/dbmanager added to handle periodic Grafana folder updates
• Collect SSD wear level metrics from newer SANtricity OS with at least one flash disk (see the main README.md how to use it in Grafana), tested with SANtricity management OS 11.52 and 11.74

Fixes

• #2 - code related to Grafana drop-down menues moved from Collector container to dbmanager so that each container doesn't drop arrays it's not aware of. The old approach worked well in upstream EPA v3.0.0 because there was only one container and thanks to the WSP it knew about all the monitored arrays
• #4 - add workarounds for different behavior of older SANtricity versions observed in v3.1-pre-release testing