Table of Contents

✨ Release Highlights

improved telemetry analysis when encountering edge-cases such as missing fields
enhance analysis when computing estimated costs per tagsRearchitected the analysis pipeline and separated telemetry analysis from cluster events/configurations loading for:
- removing the dependency of estimate cost from the telemetry so they can be computed from day one without Databricks or Cloud consumption data sources (all up tenant/account and tag distribution)
- significantly improving the scalability of the Databricks Analyzer Job (DAJ)
“all subscriptions” option on Workspace Cost page
telemetry, cost estimates and performance reporting
DLT telemetry is back! with full support for incidents and recommendations

📊 Evaluation Reports

Telemetry Snapshot

...

7684 - handle cluster idleness computation for the scenario in which the cluster has been running for several days without generating new cluster events
7470 - save the timestamp of the max cpu and max memory points of a workload/cluster
7713 - Use case-insensitive regex matching for resourceType in Azure
7771 - optimize Databricks cluster configuration retrieval
7818 - investigate missing tags in the cluster configuration
- Added better logic to retrieve the cluster config from analysis during the migration
- Added improved logic that will not reprocess the allready already processed cluster instances.
- When Databricks config not available use the latest cluster configuration.
7822 - Spot instance negative max bid price causes underestimation of vm prices

...

7653 - show cluster time brkeadown breakdown tooltip on hover over the entire cell, not only when hovering over the bar chart

...