When using a file share which resides on a NetApp filer running in 7-mode as a DataNow map point target, intermittent file operation errors may be seen.
Resumable uploads may fail, resulting in full uploads and partial temporary files left on the share in the .dataNowFolderUploads_ hidden folders on the share for 3 days before housekeeping purges them opportunistically. Downloads may transiently fail but be retried by the client. Impact to end user is likely to be low, but has the potential for increased network traffic as a result.
DataNow appliance logs may show errors such as:
ERROR [16 Nov 2016 16:25:40,605] [client-exec-4]- OrcaUserException.get(345) | ID[unknown] new OrcaUserException() from OrcaUserException.get(line:269) <=SMBFilesystemAdapter.getInputStream(line:1172) <=ResumeUtil.readTextFile(line:138)
ERROR [16 Nov 2016 16:25:40,605] [client-exec-4]- Reply.autoLog(67) | Reply:false:caught exception:SMBException [errorCode = 0xc00000e5, error=STATUS_UNSUCCESSFUL, detail=The requested operation was unsuccessful,]
ERROR [16 Nov 2016 16:25:40,605] [client-exec-4]- ResumeServiceDelegate.getMeta(546) | RESUMABLEUPLOAD
com.appsense.util.OrcaUserException: SMBException [errorCode = 0xc00000e5, error=STATUS_UNSUCCESSFUL, detail=The requested operation was unsuccessful,]
A TCPDump from the appliance will show the filer return:
STATUS_INTERNAL_ERROR;Close Response, Error:STATUS_OBJECT_NAME_NOT_FOUND
This is a known issue with NetApp Ontap (their ref 989246) and occurs if an SMB2 compound call is made to a file that is already open, if the already open instance has an oplock or lease on the object being opened.
The issue does not affect clustered Ontap since a non-compound response would be returned by the filer in this scenario.