Mailinglist Archive: opensuse (527 mails)

< Previous Next >
Re: [opensuse] A puzzle for the weekend
  • From: Per Jessen <per@xxxxxxxxxxxx>
  • Date: Sat, 13 Aug 2016 10:22:52 +0200
  • Message-id: <nomlcs$tl5$1@saturn.local.net>
Ruediger Meier wrote:

On Friday 12 August 2016, Per Jessen wrote:
I have a cron job running every 20 minutes monitoring a temperature
sensor for minor changes in trend up/down. This has been running for
a few months.

Today at 1521 I got a warning that the trend dad changed to
slightly upward. The next one failed on a DNS issue.

The puzzle:

wget -nd http://office20.local.net:4444/--slope--/1.2.3.4/0
--2016-08-12 20:38:24--
http://office20.local.net:4444/--slope--/1.2.3.4/0 Resolving
office20.local.net (office20.local.net)... failed: Name or service
not known. wget: unable to resolve host address ‘office20.local.net’

So clearly "office20.local.net" does not resolve. Except:
host office20.local.net

office20.local.net has address 192.168.3.20

ping office20.local.net

PING office20.local.net (192.168.3.20) 56(84) bytes of data.
64 bytes from office20.local.net (192.168.3.20): icmp_seq=1 ttl=64
time=0.029 ms 64 bytes from office20.local.net (192.168.3.20):
icmp_seq=2 ttl=64 time=0.040 ms

dig office20.local.net a

; <<>> DiG 9.9.4-rpz2.13269.14-P2 <<>> office20.local.net a
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 30802
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 1, ADDITIONAL:
3

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;office20.local.net. IN A

;; ANSWER SECTION:
office20.local.net. 86400 IN A 192.168.3.20


So clearly "office20.local.net" _does_ resolve. Except in wget.


No changes were applied to this system for at least a month, likely
longer.

Why can't wget resolve the address when others have no problem?


office20.local.net

Do you really own 'local.net'?

Hehe, nono. That one is only used locally on 192.168.0.0/16.

Some programs may do the lookup manually without glibc,
ignoring /etc/resolv.conf. I know that postfix does that by default.

Are you sure? At least it couldn't do it without /etc/resolv.conf, and
I'm pretty certain my installations also use /etc/hosts (which is handy
for overrides).

You should better not use domains which you don't own! Or at least a
top level domain which does not exist (hard to find one nowadays.)

I agree completely, but 'local.net' has been used internally for maybe
15 years, and changing it to something truly local would mean a lot of
checking and double-checking.

Could be that you are using another unwanted nameserver which you have
not configured to serve your own local.net. (autoconfig/dhcp),

Check, /etc/resolv.conf is fine.

Some other possibilites:
- nscd issues?

I did think of that one, and did "nscd -i hosts" - however, ping will
also use nscd, and that works.

- Maybe wget handles ipv4/v6 stupidly? Try wget -4.

Ah, forgot about that one. Yes! that was it.

What remains of the puzzle - why did this suddenly start yesterday at
15:21 after running fine since April? I swear that machine was not
touched. I don't really want to add that '-4' as it shouldn't be
necessary. Anyway, thanks for mentioning that. I'll have to look a
little closer at wget.


--
Per Jessen, Zürich (17.1°C)
http://www.hostsuisse.com/ - dedicated server rental in Switzerland.

--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
To contact the owner, e-mail: opensuse+owner@xxxxxxxxxxxx

< Previous Next >