Source of older package version?

[John McNulty]

I don't think that's it.  The jump in FORMERR count comes from
bind_reoslver_response_errrors_total but the FORMERR count from
bind_responses_total (which I assume is a count of upstream responses)
remains the same at zero or close to zero.

The Stork Agent version had previously been updated to 2.2.1 and was
running with 9.20.7, that's not responsible for the change.


On Mon, 3 Nov 2025 at 14:16, Mark Andrews <marka at isc.org> wrote:

> The security fixes cause named to need to lookup nameserver addresses
> where it didn’t before so it is talking to more servers than it used to.
> Presumably it’s just seeing more badness.
> --
> Mark Andrews
>
> On 3 Nov 2025, at 09:09, John McNulty <johnmcn1 at gmail.com> wrote:
>
> 
> Oops.  CVE not CVN.
>
> On Mon, 3 Nov 2025 at 13:57, John McNulty <johnmcn1 at gmail.com> wrote:
>
>>
>> Anyone know of a repo that retains older package releases?
>>
>> I started updating nameservers in one DC from 9.20.7 to 9.20.15 (using a
>> mirror of the copr repo) to pick up the latest CVN fixes, but my post
>> update tests highlighted a rather alarming jump in the number of FORMERR
>> and EDNS0 errors.  An increase by an order of magnitude.
>>
>> I'd like to step back through the versions to find out which release
>> after 9.20.7 introduced the change in behaviour before filing a bug
>> report.  But I've not been successful in locating a source that contains
>> them.  Copr only holds the latest release.
>>
>> Regards,
>>
>> John
>>
> --
> Visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe
> from this list.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/bind-users/attachments/20251103/eaad0d7c/attachment-0001.htm>