Downstream bug: check_dns has strange output if domain is not found

[smooge]

This was opened in Red Hat Bugzilla https://bugzilla.redhat.com/show_bug.cgi?id=1620274 but looks better suited for an upstream bug

Description of problem:
check_dns has strange and unexpected output if domain is not found

Version-Release number of selected component (if applicable):
rpm -qf /usr/lib64/nagios/plugins/check_dns
nagios-plugins-dns-2.2.1-14.20180725git3429dad.el7.x86_64
(epel-testing)

How reproducible:
always

Steps to Reproduce:
/usr/lib64/nagios/plugins/check_dns -H 1.1.1.1.sbl.spamhaus.org -q A -s 1.1.1.1

Actual results:
DNS WARNING - 1.1.1.1

Expected results:
a message containing somehow the NXDOMAIN result

Additional info:

related strace shows that original information is received:

[pid 12394] write(1, "Server:\t\t1.1.1.1\nAddress:\t1.1.1.1#53\n\n** server can't find 88.217.158.108.sbl.spamhaus.org: NXDOMAIN\n\n", 102) = 102
[pid 12393] <... read resumed> "Server:\t\t1.1.1.1\nAddress:\t1.1.1.1#53\n\n** server can't find 88.217.158.108.sbl.spamhaus.org: NXDOMAIN\n\n", 4096) = 102

[jalbstmeijer]

I'd like to confirm this unwanted behavior. Not only the output seems incomplete, also I would expect a critical state;

v1.4.15 works as expected:

check_dns -H www.google.com -s 127.0.0.1; echo $?
DNS OK: 0.010 seconds response time. www.google.com returns 172.217.19.196|time=0.009652s;;;0.000000
0

check_dns -H XXX.google.com -s 127.0.0.1; echo $?
Domain XXX.google.com was not found by the server
2

but v2.2.1 not;

check_dns -H www.google.com -s 127.0.0.1; echo $?
DNS OK: 0.116 seconds response time. www.google.com returns 74.125.193.103,74.125.193.104,74.125.193.105,74.125.193.106,74.125.193.147,74.125.193.99|time=0.116325s;;;0.000000
0

check_dns -H XXX.google.com -s 127.0.0.1; echo $?
DNS WARNING - 127.0.0.1
1

[jalbstmeijer]

Issue does not seem check_dns version related, but related to the underlying nslookup and the fact that the newer one exits with other exit codes besides 0.

version part of bind-utils-9.8.2:
system1~]# /usr/bin/nslookup -sil XXX.google 127.0.0.1; echo $?
Server: 127.0.0.1
Address: 127.0.0.1#53

** server can't find XXX.google.com: NXDOMAIN

1

part of bind-utils-9.3.6:
system2~]# /usr/bin/nslookup -sil XXX.google.com 127.0.0.1; echo $?
Server: 127.0.0.1
Address: 127.0.0.1#53

** server can't find XXX.google.com: NXDOMAIN

0

Fix could be;

--- plugins/check_dns.c.back 2018-10-19 09:18:17.402795124 -0400
+++ plugins/check_dns.c 2018-10-19 09:39:58.465156153 -0400
@@ -135,7 +135,7 @@
printf ("%s\n", command_line);

/* run the command */

• if((np_runcmd(command_line, &chld_out, &chld_err, 0)) != 0) {

• if((np_runcmd(command_line, &chld_out, &chld_err, 0)) != 0 && (np_runcmd(command_line, &chld_out, &chld_err, 0)) != 1) {
  msg = strdup(_("nslookup returned an error status"));
  result = STATE_WARNING;
  }

[sawolf]

This seems to be fixed by #494. If either of you find that the issue isn't fixed in maint, let me know and I'll reopen this issue.