Can bind be configured to not drop RR's from the cache when the upstream DNS server is unresponsive

[Barry Margolin]

In article <mailman.384.1458255932.73610.bind-users at lists.isc.org>,
 Mark Andrews <marka at isc.org> wrote:

> How do you actually expect this to ever work in real life?

I'm pretty sure Google DNS does this. Other resolver operators often get 
complaints about "Why can't I look up <whatever> through your DNS 
servers when I can do it through Google DNS?"

> Caches will generally have expired / not learnt the records by the
> time you realise that you want to keep records longer so there is
> no point even coding support for this into caches.  We don't have
> time machines.

Of course, if the record hasn't been cached in the first place, there's 
nothing you can do. But a heavily-used resolver will quickly cache most 
popular records.

When a cached record expires, the server should try to refresh it. If it 
gets a valid response, it updates the cache. But providing the old 
record if there's no response is not an unreasonable approach to fault 
tolerance.

It would be reasonable to have a configured maximum lifetime for these 
expired records, so that caches wouldn't fill up with lots of detritus 
from abandoned domains. A day seems like long enough for the 
authoritative server operator to fix their problem.

-- 
Barry Margolin
Arlington, MA