At work I got a ticket escalated to me. Resolving an hostname to an A
record stopped working.

Look me ages to figure this one out and to gut a long story short, Bind
9.16 (and maybe prior versions) hard limit the number of A record
results to 100. Someone decided it would be a good idea to make
additional A records to bring the number to 107.

It seems things were changed due to some sort of CVE that stops DNS
returning those results unless you disable the check in the named.conf
file.

I hate it when things change but nobody actually makes it obvious what
was changed...time I'll never get back.

Anyone else pissed off by Linux recently?
--
End Of The Line BBS - Plano, TX
telnet endofthelinebbs.com 23
--- SBBSecho 3.37-Linux
* Origin: End Of The Line BBS - telnet endofthelinebbs.com (86:200/30)

nelgin wrote to NixNet.NIX_LINUX <=-

At work I got a ticket escalated to me. Resolving an hostname to an A record stopped working.

Look me ages to figure this one out and to gut a long story short, Bind 9.16 (and maybe prior versions) hard limit the number of A record
results to 100. Someone decided it would be a good idea to make
additional A records to bring the number to 107.

It seems things were changed due to some sort of CVE that stops DNS returning those results unless you disable the check in the named.conf file.

I hate it when things change but nobody actually makes it obvious what
was changed...time I'll never get back.

Anyone else pissed off by Linux recently?

Was this a "Linux" thing, or a <Distro-name> thing? I'm thinking it may
be a customized version/config that's causing that...



... Gone crazy, be back later, please leave message.
=== MultiMail/Linux v0.52
--- SBBSecho 3.37-Linux
* Origin: Palantir * palantirbbs.ddns.net * Pensacola, FL * (86:200/23)

On Fri, 10 Jul 2026 08:12:00 -0500
"Gamgee" (86:200/23) <Gamgee@f23.n200.z86.fidonet> wrote:

nelgin wrote to NixNet.NIX_LINUX <=-

At work I got a ticket escalated to me. Resolving an hostname to
an A record stopped working.

Look me ages to figure this one out and to gut a long story
short, Bind 9.16 (and maybe prior versions) hard limit the
number of A record results to 100. Someone decided it would be a
good idea to make additional A records to bring the number to
107.

It seems things were changed due to some sort of CVE that stops
DNS returning those results unless you disable the check in the named.conf file.

I hate it when things change but nobody actually makes it
obvious what was changed...time I'll never get back.

Anyone else pissed off by Linux recently?

Was this a "Linux" thing, or a <Distro-name> thing? I'm thinking it
may be a customized version/config that's causing that...

It's a bind thing, which runs on linux.
--
End Of The Line BBS - Plano, TX
telnet endofthelinebbs.com 23
--- SBBSecho 3.37-Linux
* Origin: End Of The Line BBS - telnet endofthelinebbs.com (86:200/30)

nelgin wrote to Gamgee <=-

At work I got a ticket escalated to me. Resolving an hostname to
an A record stopped working.

Look me ages to figure this one out and to gut a long story
short, Bind 9.16 (and maybe prior versions) hard limit the
number of A record results to 100. Someone decided it would be a
good idea to make additional A records to bring the number to
107.

It seems things were changed due to some sort of CVE that stops
DNS returning those results unless you disable the check in the named.conf file.

I hate it when things change but nobody actually makes it
obvious what was changed...time I'll never get back.

Anyone else pissed off by Linux recently?

Was this a "Linux" thing, or a <Distro-name> thing? I'm thinking it
may be a customized version/config that's causing that...

It's a bind thing, which runs on linux.

My bad, I thought you knew more about this...

So, do you think all distros run the same version of bind? Do you think
there could *possibly* be different default configuration files
(named.conf) between different Linux distros?

Another thing - I just checked some of my various systems using
different Linux distros, and ALL of them were newer than 9.16, so maybe
that behavior has been fixed/changed? Even a Ubuntu machine had 9.18.x
and others (Slackware, Mint, MX) were 9.20.x ...

Maybe it's time to update...


... He does the work of 3 Men...Moe, Larry & Curly
=== MultiMail/Linux v0.52
--- SBBSecho 3.37-Linux
* Origin: Palantir * palantirbbs.ddns.net * Pensacola, FL * (86:200/23)

nelgin wrote to Gamgee <=-

It's a bind thing, which runs on linux.

This limit in bind was set by the ISC itself in all flavors of bind, whether for Linux, Windows, macOS, BSD, et al. So it's not just a "Linux thing".

However, it looks like they might be bringing that feature back? https://kb.isc.org/docs/rrset-limits-in-zones

It took me about 45 seconds to look this up.

-- Sean

... Knowledge is the food of the soul.

--- MultiMail/Linux
* Origin: Outpost BBS * Johnson City, TN (86:100/1)

Gamgee wrote to nelgin <=-

My bad, I thought you knew more about this...

He should being an IT professional. I mean, I didn't know about this and it took me less than a minute to ask Google about this.

Maybe it's time to update...

It's much easier to blame something than to take time time to keep up on updates. It's not a "Linux thing", it's an "ISC thing".

-- Sean

... Learn from others' mistakes; don't have others learn from you.

--- MultiMail/Linux
* Origin: Outpost BBS * Johnson City, TN (86:100/1)

Re: Re: Things that piss you off...
By: Sean Dennis to nelgin on Fri Jul 10 2026 21:22:12

It's a bind thing, which runs on linux.

This limit in bind was set by the ISC itself in all flavors of bind, whether for Linux, Windows, macOS, BSD, et al. So it's not just a "Linux thing".

I never said it didn't run on other operating systems. I said it runs on Linux

It took me about 45 seconds to look this up.

It did because I told you this existed. Maybe if all you had was "dns stopped working" like I did, it would have taken a little bit longer.
--- SBBSecho 3.37-Linux
* Origin: End Of The Line BBS - telnet endofthelinebbs.com (86:200/30)

Re: Re: Things that piss you off...
By: Sean Dennis to Gamgee on Fri Jul 10 2026 21:29:17

He should being an IT professional. I mean, I didn't know about this and it took me less than a minute to ask Google about this.

Ah, service is back to normal. I'm sorry I posted.
--- SBBSecho 3.37-Linux
* Origin: End Of The Line BBS - telnet endofthelinebbs.com (86:200/30)

It did because I told you this existed. Maybe if all you had was "dns stopped working" like I did, it would have taken a little bit longer.

I thought maybe this had to do with the "rustifying" of things. Glad to hear it was apparently something else.

I have not been frustrated lately, knock-on-wood, by linux specifically but have run into some issues where the error message from software in general has been vague. Think "file not found" but no indication as to *which* file isn't found. :( In most languages I am familiar with, that is one that is pretty easy to code for and give a proper error message.

Mike

--- ScorpioWeb v0.36a (Linux/x86_64)
* Origin: capitolcityonlinet.net * KY, USA (86:100/18)

I have not been frustrated lately, knock-on-wood, by linux specifically
but have run into some issues where the error message from software in general has been vague. Think "file not found" but no indication as to *which* file isn't found. :( In most languages I am familiar with,
that is one that is pretty easy to code for and give a proper
error message.

Yeah, we who do open source are not always great at writing proper error messages. It's like, "naaaah I'll be fine, I'll just run it through strace
or gdb if it complains". We can do better and we should :D. (It's not
really those extra 2 lines of code that provide us the dopamine though, I
have to admit...)

:)

Cheers

Peter



--- SklaffKOM v1.33rc1
* Origin: Twilight Node, Solvesborg, Sweden (86:546/20)

Hello Peter,

02 Aug 26 03:15, you wrote to All:

Yeah, we who do open source are not always great at writing proper
error messages. It's like, "naaaah I'll be fine, I'll just run it
through strace or gdb if it complains". We can do better and we should
:D. (It's not really those extra 2 lines of code that provide us the dopamine though, I have to admit...)

My BBS doors and utilities tell you what file there's an issue with and what error (my error handling unit covers errors for three different Pascal compilers; I try to be thorough) to allow for easy debugging if there's an issue.

I spent a good bit of time on that unit but it was well worth it in the end.

-- Sean

... I hear the word "icy" is easy to spell. Looking at it, now I see why.
--- GoldED+/LNX 1.1.5-b20240209
* Origin: Outpost BBS * Johnson City, TN (86:100/1)