-
Notifications
You must be signed in to change notification settings - Fork 25
Open
Description
Release version
AzureMigrateExplore-v4.0.25280.1
Describe the bug
When running a csv import-based express workflow assessment using Azure Migrate Explore (AME), the tool fails to generate performance-based outputs (Perf worksheets and Financial_Summary) due to a parsing error. This results in incomplete XLSX reports and missing visuals in the Power BI dashboard.
Expected behavior
AME should gracefully handle null values in confidenceRatingInPercentage or provide fallback logic to avoid breaking the assessment pipeline.
Actual behavior
- Perf-based sizing is skipped entirely.
- Financial modelling is not generated.
- Power BI visuals relying on these datasets are blank
Reproduction Steps
- Upload a CSV to Azure Migrate containing mandatory fields and additional performance data (CPU, memory utilization).
- Use AME to run an import-based express workflow assessment targeting IaaS, PaaS, and AVS.
- Select “Minimize time with Azure VM” as the migration strategy.
- Observe that the generated AzureMigrate_Assessment_Core_Report.xlsx lacks data in all _Perf worksheets and the Financial_Summary tab.
- Review AME logs and find the following error and warnings:
ERROR Azure VM assessment parsing error Error converting value {null} to type 'System.Double'. Path 'value[0].properties.confidenceRatingInPercentage', line 1, position 2566.
WARN Not creating excel model for All_VM_IaaS_Server_Rehost_Perf as Azure VM Perf dataset is empty
WARN Encountered an issue while processing All_VM_IaaS_Server_Rehost_Perf excel model
WARN Not generating excel model for VM_IaaS_Server_Rehost_Perf as Azure VM Perf dataset is empty
WARN Encountered issue while generating excel model for VM_IaaS_Server_Rehost_Perf
Environment
- AME version: AzureMigrateExplore-v4.0.25280.1
- Assessment type: Import-based express workflow
- Migration strategy: Minimize time with Azure VM
- Input: CSV with performance data included (CPU and Memory Utilization)
Metadata
Metadata
Assignees
Labels
No labels