Re: assertion failed: ipcache.cc:995: "tmpbuf" in 3.HEAD-CVS

ons 2008-01-09 klockan 16:28 +1300 skrev Amos Jeffries:
> In those cases its a broken server as you point out, and will respond
> with naked CNAME whether asked for A or AAAA.

If your DNS resolver server responds with a naked CNAME in response to a
A query then the requested node has a CNAME, but the node the CNAME
points at do not have an A record.

Same for AAAA queries when there is no AAAA records for the requested
host.

It's the same as a blank response. Node fount, but requested record type
not found in the node, same as when you ask for an A record on a host
only having AAAA or the other way around, without CNAME..

It's not an error response as the node as such do exist. And it's not an
empty response as there is data which may be meaningful to the requestor
even if it's not of the record type requested.

There may be other such related records in the response, for example
PTR. You should only look for and count the record you asked for, it's
safe to ignore all else.

> Ah, I get your point though.
>
> It clashes with my experimental results which made me start on the CNAME
> effort in the first place. But saying that, the failover code has had
> much improvement since then which may have skewed the results.

> I'm going to wrap all the CNAME code inside a new option
> --with-dns-cname and add a stat counter to keep track of exactly how
> many pure-CNAMEs are encountered by squid.

bare CNAMEs is expected. Nothing strange with that. See above.

> For now the option will be disabled by default, but the counter will
> always run.

Ok.

I assume there is also counters for A / AAAA queries and responses.

Regards
Henrik