Oracle Database Discussions

Hello, We have cross-coast (east and west) sites with Oracle dataguard instances running in them. The network latency between sites is about 73milli sec., We see frequent archiver errors ORA-3135, ORA-16041 etc, when the primary is trying to copy REDO to the DR site during increased REDO activity. We are using real-time apply with LGWR ASYNC method, and have dataguard broker in place for quick switchover and switchbacks. The real-time apply switches to ARCH whenever there is more activity. There is one thins that is wierd that i have seen sometimes during the REDO shipping between the primary and the standby, is that there are times when a few logs go missing during high activity, and the ARCH goes back and tries to ship the missing log, but it kinds of spins on those logs for hours. I have seen it copy the same log multiple times, but not applying it. It eventually copies it (after several attempts), but this causes our databases to stay out of sync for long durations. Has anyone seen such behavior and the errors mentioned above in their dataguard environment? Are there any network/firewall/router con Please help.