All,
We had a discussion this morning during the hackathon about a value with
the document caching-resolution-failures. The current text in 3.2 says:
Resolvers MUST cache resolution failures for at least 5 seconds. The
value of 5 seconds is chosen as a reasonable amount of time that an
end user could be expected to wait.
Resolvers SHOULD employ an exponential backoff algorithm to increase
the amount of time for subsequent resolution failures. For example,
the initial time for negatively caching a resolution failure is set
to 5 seconds. The time is doubled after each retry that results in
another resolution failure. Consistent with [RFC2308], resolution
failures MUST NOT be cached for longer than 5 minutes.
There was some operational feedback that suggests 1 second is also
a very reasonable value here. With some discussion, here is some
suggested text:
Resolvers MUST cache resolution failures for at least 1 second.
The initial duration SHOULD be configurable by the operator. A
longer cache duration for resolution failures will reduce the
processing burden from repeated queries, but will also lengthen
the recovery period from transitory issues.
Resolvers MAY* employ an exponential backoff algorithm to increase
the cache duration when resolution failures are persistent. For
example, the initial time for negatively caching a resolution
failure could be set to 5 seconds, and doubled after each retry
that results in another resolution failure, up to a configurable
maximum.
Consistent with [RFC2308], resolution failures MUST NOT be cached
for longer than 5 minutes.
---
* Note that the original text has this as SHOULD. I've heard reasons for
both SHOULD and MAY.
We'd like to hear from the working group on this value, and what the
working group thinks of this change
thanks
tim
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop
