Remote NDMP Restore hangs after media gets positioned.
From the job details
Feb 21, 2011 3:42:08 PM - mounted <media_id>; mount time: 0:00:38
Feb 21, 2011 3:42:09 PM - positioning <media_id> to file 5
Feb 21, 2011 3:42:28 PM - positioned <media_id>; position time: 0:00:19
Restore hangs at this point.
BPTM log (/usr/openv/netbackup/logs/bptm) to just stop immediately after attempting to read the tape information after the mount -
11:21:52.648  <4> read_backup: begin reading backup id <backup_id> (restore), copy 1, fragment 1, from media id <media_id> on drive <drive_name> (index #)
11:21:52.648  <2> io_close: closing /usr/openv/netbackup/db/media/tpreq/drive_<drive_name>, from bptm.c.15923
11:21:52.651  <2> process_tapealert: TapeAlert returned 0x00000000 0x00000000 (from tpunmount)
NetBackup 7.0.1 introduces the ability to perform multiplex backup of remote NDMP servers.
The 6.5.x media server is not capable of performing a multiplex NDMP restore.
In the bprd log (/usr/openv/netbackup/logs/bprd) determined that there was a media server at NetBackup 6.5.2 level in this environment -
12:00:40.960  <2> local_getHostInfo: platform of client 2ksewrwww1.tak.bup is Linux, RedHat2.6
12:00:41.030  <2> bpcr_get_uname_rqst: bpcr.c: 3777 count=22, buffer=UNAME_SYSNAME = Linux
12:00:41.030  <2> bpcr_get_uname_rqst: bpcr.c: 3777 count=43, buffer=UNAME_NODENAME = 2ksewrwww1.taketwo.online
12:00:41.310  <2> getPatchLevel: level = >6.5.2<
Per TechNote TECH145842 the workaround was to create NON_MPX_RESTORE touch file on the master server. This did not work in this case.
Only applying engineering binary for ET2245820 resolved the problem.
NetBackup 7.0.1 Master Servers doing NDMP restores to NetBackup 7.0 and lower media servers.