Summary
NDMP backups using Clustered Data ONTAP 8.3 as a destination, which has been configured with the Cluster Aware Backup (CAB) extension (tape drives are controlled by Clustered Data ONTAP), may fail with a Drive fail" error message shortly after mounting the tape."
Symptoms
A typical error message sequence in the job log is:
... Node(xx.xx.xx.xx) drive(tape70-G01) volser(B00045) partition(1) mount rc=(0)
... NDMP server entry: Type(NDMP_LOG_NORMAL), ID(1), Text(Session identifier: 47792 )
... *** Drive fail reported: drive tape70-G01 on node W2K8R2 in job 1429998180 ***
... *** Fail searching socket 5 of module ssdmbs at node [dev ] ***
The SSNDMPC log file (.ncl suffix) on the client/proxy node for the storage system will show a NDMP Not Supported error, in the same timeframe as the "Drive fail" error:
... ERROR: Task 1456, type(BACKUP), state(CONFIG_18); Error(100, NC_TASK_ERROR_NDMP_RECV_MSG), NDMP error(1, NDMP_NOT_SUPPORTED_ERR); Connection 2572, Message token 542244867
... Task 1456, type(BACKUP), state(CONFIG_18); Unrecoverable error.
Resolution
NetApp is currently working to resolve the issue. Please refer to BURT 898836 for more information on the issue. Until NetApp announces the fix, the following workarounds are possible to enable NDMP backup for Clustered Data ONTAP (CDOT):
- Use Node-Scope operation mode allowing configuration and use of tape drive directly to the controller. Node-scope enabling and description can be found in Data ONTAP Cluster Mode Installation and Support.
- NetApp controller can be configured in a Cluster operation mode and backup is configured to run to the device server (a non-CDOT NetApp controller or Windows or Linux server which is a part of the DPX Enterprise).