Couchbase "issue"...logs rolled over in 7 hours

Almost all of the log files in this example have rolled over in 7 hours. This is just way too fast for support to be able to do anything. While this time missed the event by a mere 2 hours, in many cases it will be over the weekend or a few days before the customer actually sends over logs. At this point, we can do nothing more than say "sorry, we don't know what happened", ask the customer to suffer another outage and be quicker next time.

There has got to be something we can do to improve this core part of the supportability of the product:
-Allow the user to set how much space they want to set aside of logs, above the current 1GB threshold
-Compress our rolled_over log files
-Allow for exporting to syslog
-Further improvement of the logging system to not "spam" the logs with repeated messages