We had a server die on us, so we brought a new one up with the same name/IP. When we go to backup now, full or incremental, we get a socket write error. I checked and the new client's nic is set to 100mb full duplex. It will run anywhere from 10mins to 2 hours before the error happens. I checked the timeout and buffer settings and raised them.
Has anyone encountered this socket error?
Client is a Win2008 64bit
Master/media is Win2003 64bit
NetBackup 7 using the media dedup
10/29/2010 1:37:35 PM - estimated 1790412 Kbytes needed
10/29/2010 1:37:35 PM - started
10/29/2010 1:37:36 PM - started process bpbrm (10124)
10/29/2010 1:37:43 PM - connecting
10/29/2010 1:37:46 PM - connected; connect time: 00:00:03
10/29/2010 1:37:56 PM - begin writing
10/29/2010 1:38:21 PM - Warning bpbrm(pid=10124) from client msgsvmodrive: WRN - old TIR info file 'C:\Program Files\Veritas\NetBackup\tir_info\E\88MDG\NetBackup_file_info.O_Drive' is missing. Backing up everything in 'E:\88MDG'
10/29/2010 1:51:38 PM - Critical bpbrm(pid=10124) from client msgsvmodrive: FTL - socket write failed
10/29/2010 1:51:51 PM - Info msgsapfyb5(pid=956) StorageServer=PureDisk:msgsapfyb5; Report=PDDO Stats for (msgsapfyb5): scanned: 3889376 KB, stream rate: 19.59 MB/sec, CR sent: 310551 KB, dedup: 92.0%, cache hits: 23963 (70.1%)
10/29/2010 1:51:54 PM - end writing; write time: 00:13:58
socket write failed(24)
