> I asked a similar question 2 weeks ago and got a non-response (e.g., a > response with no real information).
The only somewhat good reason I see to overriding (well, lowering) the cache time is if it causes your server any memory issues. Although the real solution then would be to buy more memory. Yes, an active DNS server will cache a few GB, depending on usage patterns, how common DNSSEC becomes etc, but if you run an active DNS-server I'd hope you'd be able to get the budget for that memory. Overriding the cache TTL by lowering it is essentially the same as what happens when nameservers are restarted - it isn't optimal, but it happens all the time all over the world. Overriding the cache TTL by _increasing_ the value is something that's bound to break many setups - if I set my TTL to a low value, it's hopefully for a reason. I have had to remove some cached information before it timed out by itself due to TTL - depending on how often you need to do that and how many servers you have, one option might be to do something like "rndc flushname hostname.to.flush" on those servers. Depending on your setup, you might also consider centralizing this so you can do it once from one location (easiest solution: make a wrapper script, running rndc on all servers in turn, over the network). Regards Eivind Olsen _______________________________________________ bind-users mailing list bind-users@lists.isc.org https://lists.isc.org/mailman/listinfo/bind-users
