Problem
Backup fails intermittently with status code 24: socket write failed.
Error Message
Snippet of bpbkar log:
06:11:47.026 [25723] <16> bpbkar: ERR - Cannot write to STDOUT. Errno = 32: Broken pipe
06:11:47.026 [25723] <16> bpbkar: ERR - bpbkar FATAL exit status = 24: socket write failed
06:11:47.026 [25723] <4> bpbkar: INF - EXIT STATUS 24: socket write failed
HINT: bpbkar log can be found in /usr/openv/netbackup/logs/bpbkar/ (UNIX/Linux client) or <install path>\Veritas\NetBackup\logs\bpbkar\ (Windows client)
Cause
Depending on the amount of data being backed up, the "TCP keep alive interval" may need to be shortened so NetBackup media server is aware that the client is still actively doing backup.
Solution
To verify the TCP keep alive interval value on Solaris 9 / 10 client, run:
/usr/sbin/ndd -get /dev/tcp tcp_keepalive_interval
By default, the value is 7200000 (2 hours). Recommended value for large backup is 300000 (5 minutes).
Simply run the following command to set:
/usr/sbin/ndd -set /dev/tcp tcp_keepalive_interval 300000
Applies To
Solaris 9 / 10 OS + NetBackup 6.5.x.