Azure Batch updates for May 2017
發佈日期： 五月 24, 2017
New and important Azure Batch capabilities are now available:
- Low-priority VMs to enable compute at a fraction of the price (preview):
- Low-priority VMs are allocated from surplus compute capacity and are available for up to an 80% discount, enabling certain types of workloads to run for a significantly reduced cost or allowing you to do much more for the same cost.
- For more information, see the announcement, Batch documentation, and pricing details.
- Metered-based licensing for graphics and rendering software with Azure Batch Rendering (private preview):
- Support for built-in output file upload from pool nodes to Azure Blob storage:
- Batch has always provided the built-in ability to download files by copying them from Azure Blob storage to the pool nodes via task resource files. Now there is built-in support to persist task output files to Blob storage. You can specify output files on a task.
- Output file support is available only for pools that you create by using virtual machine configuration (not cloud service configuration).
- For more information, see the Add a task to a job article.
- Improved support for creating Linux VM users:
- When you're specifying the user accounts to be created on each node in pool configuration, you can specify a UID and GID for Linux users.
- For more information, see the Add a pool to an account article.
- Improved task status reporting:
- The execution info for a task now has a result property that indicates if the task succeeded or failed.
- For more information, see the Get information about a task article.