Summary
Instant Access (IA) mapping fails because of excessive NetApp Data Fabric Manager (DFM) use of the http connection, RC 23001. This article provides recommended actions to work around the NetApp issue.
Symptoms
IA mapping of restore points from block-level backup fails with RC 23001.
Mapper log (*.mpr) on the node to which the mapping is attempted displays the following:
SNBMPR3066D 7440: check_if_lun_already_exist: lun_list_info() failed
SNBMPR3069E 7440: create_and_map_lun(): Failed to cleanup old temp Luns
SNBMPR3066D 7440: create_and_map_lun(): starting Step2
SNBMPR3069E 7440: HTTP POST - failed, status code = 507
SNBMPR3069E 7440: create_and_map_lun() failed. rt=23001
SNBMPR3069E 7440: Failed to create Lun on secondary(Netapp Storage System name)
SNBMPR3066D 7440: createAndMapLun() failed. rt=23001
Resolution
NetApp opened a bug for this DFM issue. The following sources provide information:
To work around the NetApp issue as it relates to IA map, try increasing the http maximum connection limit from its default value of 512 to the maximum value of 1023. Use the following commands as needed:
- To check the current value for httpd.admin.max_connections, enter
options http
- To increase the maximum connection limit from 512 to 1023, enter
options httpd.admin.max_connections 1023
- To display http statistics on the storage system, enter
httpstats -a
If the peak connections are equal to the maximum connection limit, shut down DFM temporarily until the IA map and its associated restore are completed.