BIND 10 #859: Search cache for lowest reachable delegation
BIND 10 Development
do-not-reply at isc.org
Fri Apr 22 20:16:10 UTC 2011
#859: Search cache for lowest reachable delegation
-------------------------------------+-------------------------------------
Reporter: | Owner: UnAssigned
vorner | Status: reviewing
Type: | Milestone:
defect | Sprint-20110503
Priority: major | Resolution:
Component: | Sensitive: 0
Unclassified | Sub-Project: DNS
Keywords: | Estimated Difficulty: 0.0
Defect Severity: N/A | Total Hours: 0
Feature Depending on Ticket: |
Add Hours to Ticket: 0 |
Internal?: 0 |
-------------------------------------+-------------------------------------
Changes (by vorner):
* owner: vorner => UnAssigned
* status: new => reviewing
Comment:
Hello
The code is ready. It took me some time to think it over, but it turned
out to be quite simple in the end.
I decided I don't want to take the lowest NS in cache unconditionally.
Because if I did, the A/AAAA records for it might have expired. The NSAS
would get the NS from cache, asked for IP addresses, which would generate
query to the same zone, and we would get a loop. So I check that the NS
has some address, if not, I move up to the next known delegation. This
should not increase the number of queries, because we should get the
A/AAAA with the delegation as glue anyway.
Unluckily, I can't test it in reality, administrator of my network blocks
port 53 except his own nameserver. So, if someone could have a look if
this works and if it solves #833? (It should, in theory, I counted that we
could have quadratic number of queries in the number of delegations
before).
I propose this changelog entry:
{{{
[func] vorner
We use cache to find lowest usable delegation point, reducing number of
sent queries.
}}}
--
Ticket URL: <http://bind10.isc.org/ticket/859#comment:1>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list