Certain features have been removed or deprecated in StorageGRID Webscale 11.1. You must review these items to understand whether you need to update client applications or modify your configuration before you upgrade.
NetApp-provided virtual machine disk files and scripts for OpenStack are no longer supported for StorageGRID Webscale installation, expansion, or recovery operations. If you need to recover a node running in an OpenStack deployment, download the StorageGRID Webscale 11.1 files for your Linux operating system. Then, follow the recovery and maintenance instructions for replacing a Linux node.
ILM rules that were created before StorageGRID Webscale 10.3 were deprecated in version 10.4, and are not supported in StorageGRID Webscale 11.1. Before upgrading, you must recreate any legacy rules you still want to use and then delete all of them. See "Managing legacy ILM rules" for instructions.
The consistency level used by default for new S3 buckets or Swift containers has been renamed. The previous name was "default." The new name is "read-after-new-write."
The "read-after-new-write" consistency level provides read-after-write consistency for new objects and eventual consistency for object updates. It offers high availability and data protection guarantees and matches AWS S3 consistency guarantees.
Because of this change, you might need to update your applications before upgrading to StorageGRID Webscale 11.1. To understand what you need to change, refer to the table, which compares how StorageGRID Webscale behaved in version 11.0 to how it behaves in version 11.1 for each type of request that you can use to specify consistency level.
Type of request | Behavior in version 11.0 | Behavior in version 11.1 |
---|---|---|
PUT /org/containers/{containerName}/consistency in the Tenant Management API | You could specify "default" for the consistency. | You can continue to specify "default" or you can specify "read-after-new-write" to achieve the same consistency behavior. |
GET /org/containers/{containerName}/consistency in the Tenant Management API | "default" could be returned. | If the bucket used the "default" level in version 11.0 or the "read-after-new-write" level in version 11.1, "default" is returned (for backward compatibility). Note: The "default" level is deprecated in version 2 of the Tenant Management API. In version 3 of the API (future release of StorageGRID Webscale), "default" will no longer be returned.
|
Configure Consistency Level dialog in the Tenant Manager | You could select "Default" as the consistency. |
|
|
Bucket was created with the "default" consistency control. | Bucket is created with the equivalent "read-after-new-write" consistency control. |
|
You could specify "default" as the consistency control. | You must specify "read-after-new-write" to achieve the same consistency behavior. |
|
"default" could be returned. | "read-after-new-write" is now returned instead of "default." |
S3 HEAD Object request | You could specify "default" as the consistency control. | You can specify "read-after-new-write" in the consistency control header to achieve the same consistency behavior, or you can configure the S3 bucket to use the "read-after-new-write" consistency level. |
Audit export through CIFS/Samba is deprecated in StorageGRID Webscale 11.1, and will be removed in a future StorageGRID Webscale release.
If you previously configured audit clients for CIFS, you should now configure audit clients for NFS instead. See the instructions for administering StorageGRID Webscale.
The cache_device[cache_device_config_attributes][type] is no longer needed when creating a cache device using the API.
The cache_device[cache_device_config_attributes][format] is no longer needed when creating a cache device using the API. Previously, the cache_devices API call required the cache device format to be specified. Now the format is always EXT4.
The active_directory[primary_dc] is no longer needed when creating an Active Directory using the API.