Large log files

I work on the performance of an application, and don't have much experience of production DBA activates.

However, I'd like to know what the management implications are of very large redo log files. The application on a large system generates redo at rates of well over 10MB/s.

Specifically:

At these rates, does the 30 minute between log switches rule of thumb still hold? i.e. should I be recommending redo log files of the order of several tens of gigabytes?

How would a DBA go about archiving log files that can be generated at rates greater than can be easily streamed onto a single tape?

I assume that recovery would be a problem (as in take geological time) with large redo logs. Is there any way of improving this, or minimising it (other than avoiding unplanned outages)?

In general, what other problems would be posed by such large log files?

Thanks,

-- 
Andrew Mobbs - http://www.chiark.greenend.org.uk/~andrewm/