Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
Fix automatic DNS server name detection to honor -4 and -6 switches
(default to 127.0.0.1 or ::1 respectively, i.e. if -6 is given ::1
is used instead of 127.0.0.1)
|
|
FreeBSD starting with version 10 is shipped with ldns instead of bind as
resolver. Consequently the dig tool in base is replaced by drill. While dig can
still be installed as a third party application, it would be nice to make do
with the tools available in the system already.
This patch rearranges the command line used to invoke dig slightly so that it
can be used with both dig and drill (tested with dig 9.8.3-P1 and 9.9.4 as well
as drill 1.6.16). It would be really neat if the configure script could be
changed to automatically pick up drill when dig is not available (or the other
way around), but my autotools-foo is not good enough for that.
This part of the patch is an extended version of the locally maintained patch
currently deployed in the FreeBSD ports tree by Dmitry Sivachenko.
|
|
This patch verifies if the response is returned by the dns server we
are trying to check
---
Closes #981
Closes #982
Closes #1164
|
|
Signed-off-by: Sven Nierlein <sven@nierlein.de>
|
|
Conflicts:
plugins/netutils.c
|
|
I'm submitting a small patch to check_disk that will show which partitions are
actually causing the error state. This helps the human operator to quickly
identify the exact issue. I made it so a single -v flag is required so it
doesn't harm legacy code that may be parsing the original style of output.
Thanks to Jason Benguerel.
---
Closes #984
Closes #1182
|
|
Closes #907
Closes #1184
|
|
Not sure if this is of use or not - we have a strange requirement to run
certain servers 5 minutes fast. I've added a switch to the check_ntp_time
to allow for this offset.
Thanks to Patrick McAndrew.
|
|
Changed bool to short.
Removed first instance of is_socket to avoid
redeclaration error.
Changed 'socket' to 'file socket' for verbosity.
|
|
thus simplify the problem debugging: no need to check for this information in
the Nagios configuration.
This function is only used by 'check_tcp.c'.
Without the patch:
$ ./plugins/check_tcp -H 127.0.0.1 -p 21
Connection refused
$ ./plugins/check_tcp -H /var/spool/nagios/cmd/nagios.cmd
Permission denied
With the patch:
$ ./plugins/check_tcp -H 127.0.0.1 -p 21
connect to address 127.0.0.1 and port 21: Connection refused
$ ./plugins/check_tcp -H /var/spool/nagios/cmd/nagios.cmd
connect to socket /var/spool/nagios/cmd/nagios.cmd: Permission denied
Thanks to Davide Madrisan.
---
Closes #1277
|
|
This patch allows checking if MySQL server is running without providing valid
username and password. Similar to check_ssh plugin it returns MySQL server
version string and protocol number.
Example:
check_mysql -n -H aaa.bbb.ccc.ddd
MySQL OK - Version: 5.0.51a-24+lenny5 (protocol 10)
This is useful for monitoring servers where one does not have administrator
privileges or does not want to grant any privileges for the monitoring station.
To enable this functionality new option --ignore-auth (-n) is added to
check_mysql plugin.
Thanks to Julius Kriukas
Closes #1020
Closes #1178
|
|
Adjust the expected date strings to the now-localized output produced by
plugins/sslutils.c.
Closes #1275.
|
|
Perl said: "Unrecognized escape \d passed through at ./t/check_dns.t
line 57."
|
|
sshutils prints the expiry time of certificates in US format
this patch uses the strftime %c, I don't know how portable that is
Thanks to Neil Prockter.
Closes #1188
Closes #1161
Closes #977
Closes #976
Closes #975
Closes #840
Closes #382
|
|
|
|
---
Closes #1160 and #973
|
|
|
|
|
|
closes #1236 and #1239
|
|
be easy to see which check caused the alert.
|
|
Forking raises a race condition, where the parent might run the
test before the child has had time to fork. If that happens,
an error similar to this is produced:
Failed test 'Output correct'
at ./t/check_procs.t line 32.
'PROCS OK: 0 processes with args 'sleep 7' | processes=0;;;0;'
doesn't match '/^PROCS OK: 1 process?/'
Sleeping a bit should avoid the problem. It might be enough to
sleep less than a second, but perl's built-in sleep function only
supports integer seconds.
In our build environment, the build failed 3 of 4 times before
this patch. After the patch it failed 0 of 7 times.
Signed-off-by: Mikael Falkvidd <mfalkvidd@op5.com>
|
|
On at least FreeBSD, NetBSD, OpenBSD, DragonFly, and OS X, this is
hard-coded at 104 bytes.
Closes #1267.
|
|
|
|
|
|
check_ssh: check protocol
|
|
Include common.h before any system headers.
|
|
check_nt.c - Changed 'Mb' to 'MB' in MEMUSE.
|
|
Serveral coverity fixes from nagios-plugins
|
|
Signed-off-by: Sven Nierlein <sven@nierlein.de>
|
|
Added two if elses to cover when only one threshold is set.
|
|
|
|
|
|
will help the admins when multiple checks are configured
Signed-off-by: Davide Madrisan <davide.madrisan@gmail.com>
|
|
|
|
Coverity 66479 - validate_arguments has no need to return anything, as it dies on error, yet was set to return an int. Set to void to resolve warning.
|
|
Coverity 66480 - Potential array index out of bounds, since result was not verified to be positive prior to using as an index for state[]. Simply reording the if statement should resolve the issue. - SR
|
|
Coverity 66514 - Possible leakage and overflow with addr in redirect functionality. Not confirmed as null terminated, and externally gathered. Restrict string comparisons and duplications by size. - SR
|
|
Coverity 66524 - req.data is not neccessarily null terminated but still feed to printf statements. This both does that, and verifies the struct more so than before. - SR
|
|
Recv into buffer is not properly null terminated prior to strstr and possible other string functions expecting a null termination. Simply take bytes received and use as an index to append \0 after. We are creating buffer[] with size of MAX_INPUT_BUFFER and recv with MAX_INPUT_BUFFER-1 so this should never overflow.
|
|
Coverity 66531 - ereg.buffer can be printed without being initialized if do_include and do_exclude are null and critical is an invalid regex. While minor this may leak memory and cause undefined behavior.
|
|
This should fix some problems building on AIX.
|
|
also make test is now working on travis-ci
Signed-off-by: Sven Nierlein <sven@nierlein.de>
|
|
|
|
|
|
|
|
Our die() function doesn't append a newline character to the message.
|
|
Allow for using the FreeRADIUS Client library instead of radiusclient or
radiusclient-ng. The latter two projects are dead.
Closes #1231.
|
|
The puts(3) function already appends a newline character to the string.
|
|
when returning syscontact. So make them optional since we want to test
check_snmp and not the snmpd.
Signed-off-by: Sven Nierlein <Sven.Nierlein@consol.de>
|