Master server: Solaris 8
NetBackup 4.5 FP4

We experience restores hanging up: Images are found, a list of required tapes 
is generated and then bprd log reports that it's waiting for ports to close. It 
remains in this state for hours without the ports closing. 

Killing all the bprd processes, restarting bprd and re-submitting the restore 
is the only action that gets the restores going.

The tcp close_wait interval has been reduced to 5000. (We have tried to reduce 
it to 1000, but it made no difference.)

This is happening on a daily basis and is driving the customer (and us!) crazy. 
The customer is doing daily restores at their DR site.

Any ideas??


master:/> bpps|grep bprd
    root 14530 21004  0 08:18:45 ?        0:37 /usr/openv/netbackup/bin/bprd
    root 14525 21004  0 08:18:25 ?        0:08 /usr/openv/netbackup/bin/bprd
    root 25184 21004  0 18:37:14 ?        0:06 /usr/openv/netbackup/bin/bprd
    root 23012 21004  0 18:16:41 ?        0:07 /usr/openv/netbackup/bin/bprd
    root 14855 21004  0 08:22:16 ?        0:37 /usr/openv/netbackup/bin/bprd
    root 21004     1  0 17:57:31 ?        0:01 /usr/openv/netbackup/bin/bprd
    root 21427     1  0 18:02:47 ?        0:01 bprd -dontfork -mpxmain
    root 14520 21004  0 08:18:19 ?        0:37 /usr/openv/netbackup/bin/bprd



master:/> netstat -a |grep bprd
      *.bprd               *.*                0      0 24576      0 LISTEN
master.bprd master.2745 32768      0 32768      0 CLOSE_WAIT
master.bprd master.2491 32768      0 32768      0 CLOSE_WAIT
master.bprd master.4512 32768      0 32768      0 CLOSE_WAIT
master.bprd master.3359 32768      0 32768      0 CLOSE_WAIT
master.bprd master.1928 32768      0 32768      0 CLOSE_WAIT
master.bprd master.1629 32768      0 32768      0 CLOSE_WAIT
30001d001b8 stream-ord 30006552c00 00000000 /usr/openv/var/vnetd/bprd.uds   

Marianne van den Berg