[Dnsmasq-discuss] Implications of raising dns-forward-max

[Petr Menšík]

I think these kinds of problem arises usually when connectivity to 
upstream is broken or interrupted.

150 is number of concurrent queries waiting for resolution. It should 
not be required to increase this limit if the remote server is answering 
fast. Unless I am mistaken, as soon as the response arrives back, that 
frees on of those sessions. Dnsmasq cannot drive retries itself, it 
relies on client to do them. But that should not matter.

Unless we know more details when and how increasing this value should 
help, I would be opposed in doing so. We may need to improve switching 
to alternative resolvers, if that alternatives respond faster. If that 
is interruption in upstream connectivity, no increase here would help.

On 29. 06. 23 15:39, Eric Fahlgren wrote:
> On Thu, Jun 29, 2023 at 12:52 AM Buck Horn <buckhorn at weibsvolk.org> wrote:
>
>     Blindly increasing dns-forward-max without having analysed the actual
>     issue does not seem to recommend itself as the best option
>     available to
>     me, even if it would turn out that actual resource impacts are minor.
>
>     Stopping or unconfiguring the DNS loop or switching to more reliable
>     upstreams would seem more adequate measures to address the warning.
>
>     If you have actively been involved in investigating such an issue, it
>     may be worth verifying those causes, if only to preclude them. ;)
>
>
> Hi Buck,
>
> Thanks a bunch for the possible causes.  As you could probably tell 
> from my "assuming", I have not done any root cause investigation, all 
> the reports are just anecdotal "I saw this, then did that, and the 
> issue went away..." In some of the posts, it has been suggested that 
> the installation default be raised to 500 based on these reports, 
> because "why not?" and "it looks like it fixes something?"
>
> So, at this point I'm still in tentative data collection phase and 
> armed with your suggested avenues of investigation, I'll see if I can 
> get someone to reproduce the problem and find the real root issue.
>
> Thanks again,
> Eric
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> https://lists.thekelleys.org.uk/cgi-bin/mailman/listinfo/dnsmasq-discuss

-- 
Petr Menšík
Software Engineer, RHEL
Red Hat,http://www.redhat.com/
PGP: DFCF908DB7C87E8E529925BC4931CA5B6C9FC5CB
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/attachments/20230703/ef612df3/attachment.htm>