Checkmk git commits October 2017

checkmk-commits@lists.checkmk.com

10 participants
288 discussions

Added additional test for unreachable notifications

by Lars Michelsen

Module: check_mk Branch: master Commit: 26190e8fcc8c68453e7c4543ad8edd0f77c2afda URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=26190e8fcc8c68… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Tue Oct 17 13:47:18 2017 +0200 Added additional test for unreachable notifications Change-Id: I0b6a9df498d8e86adf686fcd6620229cbd70e93b --- .../test_unreachable_notifications.py | 126 ++++++++++++++++++++- 1 file changed, 125 insertions(+), 1 deletion(-) diff --git a/tests/notifications/test_unreachable_notifications.py b/tests/notifications/test_unreachable_notifications.py index 3d821e8..9b2740b 100644 --- a/tests/notifications/test_unreachable_notifications.py +++ b/tests/notifications/test_unreachable_notifications.py @@ -356,4 +356,128 @@ def test_child_up_after_parent_recovers(scenario, site, initial_state): # - Set parent UP, expect UP notification for child site.send_host_check_result("notify-test-parent", STATE_UP, "UP") - log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UP;check-mk-notify;") + if scenario.unreachable_enabled: + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UP;check-mk-notify;") + else: + log.check_not_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UP;check-mk-notify;") + + +# Test the situation where: +# a) Child goes down +# b) Parent goes down, child becomes unreachable +# d) Parent goes up, child becomes down +def test_down_child_becomes_unreachable_and_down_again(scenario, site, initial_state): + log = HistoryLog(scenario.core) + + # - Set child down, expect DOWN notification + site.send_host_check_result("notify-test-child", STATE_DOWN, "DOWN") + log.check_logged("HOST ALERT: notify-test-child;DOWN;HARD;1;DOWN") + + if scenario.core == "cmc": + # CMC: Send a new check result for the parent to make the CMC create the host notification + # for the child + site.send_host_check_result("notify-test-parent", STATE_UP, "UP") + + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;DOWN;check-mk-notify;") + + # - Set parent down, expect DOWN notification for parent and UNREACHABLE notification for child + site.send_host_check_result("notify-test-parent", STATE_DOWN, "DOWN") + log.check_logged("HOST ALERT: notify-test-parent;DOWN;HARD;1;DOWN") + + # Difference beween nagios/cmc: when sending DOWN via PROCESS_HOST_CHECK_RESULT + # the nagios core needs another child down check result to report it as unreachable. + if scenario.core == "cmc": + log.check_logged("HOST ALERT: notify-test-child;UNREACHABLE;HARD;1;") + + if scenario.unreachable_enabled: + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;") + # TODO: Can not test this because it drains too many entries from the log. HistoryLog could deal + # with this by readding the read lines after succeeded test or similar + #else: + # log.check_not_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;") + + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-parent;DOWN;check-mk-notify;") + + elif scenario.core == "nagios": + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-parent;DOWN;check-mk-notify;") + + site.send_host_check_result("notify-test-child", STATE_DOWN, "DOWN", expected_state=STATE_UNREACHABLE) + log.check_logged("HOST ALERT: notify-test-child;UNREACHABLE;HARD;1;") + + if scenario.unreachable_enabled: + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;") + + # - Set parent up, expect UP notification + site.send_host_check_result("notify-test-parent", STATE_UP, "UP") + log.check_logged("HOST ALERT: notify-test-parent;UP;HARD;1;") + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-parent;UP") + + # - Next child check DOWN + # cmc: expect no notification (till next parent check confirms UP) + # nagios: expect notification without + site.send_host_check_result("notify-test-child", STATE_DOWN, "DOWN") + log.check_logged("HOST ALERT: notify-test-child;DOWN;HARD;1;") + + if scenario.core == "cmc": + # - Set parent UP (again), expect DOWN notification for child + site.send_host_check_result("notify-test-parent", STATE_UP, "UP") + + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;DOWN;check-mk-notify;") + + +# Test the situation where: +# a) Child goes down +# b) Parent goes down, child becomes unreachable +# c) Child goes up +def test_down_child_becomes_unreachable_then_up(scenario, site, initial_state): + log = HistoryLog(scenario.core) + + # - Set child down, expect DOWN notification + site.send_host_check_result("notify-test-child", STATE_DOWN, "DOWN") + log.check_logged("HOST ALERT: notify-test-child;DOWN;HARD;1;DOWN") + + if scenario.core == "cmc": + # CMC: Send a new check result for the parent to make the CMC create the host notification + # for the child + site.send_host_check_result("notify-test-parent", STATE_UP, "UP") + + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;DOWN;check-mk-notify;") + + # - Set parent down, expect DOWN notification for parent and UNREACHABLE notification for child + site.send_host_check_result("notify-test-parent", STATE_DOWN, "DOWN") + log.check_logged("HOST ALERT: notify-test-parent;DOWN;HARD;1;DOWN") + + # Difference beween nagios/cmc: when sending DOWN via PROCESS_HOST_CHECK_RESULT + # the nagios core needs another child down check result to report it as unreachable. + if scenario.core == "cmc": + log.check_logged("HOST ALERT: notify-test-child;UNREACHABLE;HARD;1;") + + if scenario.unreachable_enabled: + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;") + # TODO: Can not test this because it drains too many entries from the log. HistoryLog could deal + # with this by readding the read lines after succeeded test or similar + #else: + # log.check_not_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;") + + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-parent;DOWN;check-mk-notify;") + + elif scenario.core == "nagios": + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-parent;DOWN;check-mk-notify;") + + site.send_host_check_result("notify-test-child", STATE_DOWN, "DOWN", expected_state=STATE_UNREACHABLE) + log.check_logged("HOST ALERT: notify-test-child;UNREACHABLE;HARD;1;") + + if scenario.unreachable_enabled: + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;") + + # - Set child up, expect: + # cmc: pending UP notification, and sent up notification after next parent check + # nagios: UP notification + site.send_host_check_result("notify-test-child", STATE_UP, "UP") + log.check_logged("HOST ALERT: notify-test-child;UP;HARD;1;") + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-child;UP") + + ## - Set parent up, expect UP notification + site.send_host_check_result("notify-test-parent", STATE_UP, "UP") + log.check_logged("HOST ALERT: notify-test-parent;UP;HARD;1;") + log.check_logged("HOST NOTIFICATION: check-mk-notify;notify-test-parent;UP")

6 years, 11 months

5321 FIX mk_oracle: Fixed usage of awk command in case of SunOS which causes ' record ... too long' error

by Simon Betz

Module: check_mk Branch: master Commit: 38bfd092a9197aecc6a429894e6ed886b67120ba URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=38bfd092a9197a… Author: Simon Betz <si(a)mathias-kettner.de> Date: Tue Oct 17 13:56:25 2017 +0200 5321 FIX mk_oracle: Fixed usage of awk command in case of SunOS which causes 'record ... too long' error The consequence of above error was that Check_MK did not find any ORACLE instances any more. Fixed that by using {{nawk}} in case of SunOS. Change-Id: I73612f7e8412f0b62bd7a807718bd672a32861f8 --- .werks/5321 | 13 +++++++++++++ agents/plugins/mk_oracle | 4 +++- 2 files changed, 16 insertions(+), 1 deletion(-) diff --git a/.werks/5321 b/.werks/5321 new file mode 100644 index 0000000..b1d2556 --- /dev/null +++ b/.werks/5321 @@ -0,0 +1,13 @@ +Title: mk_oracle: Fixed usage of awk command in case of SunOS which causes 'record ... too long' error +Level: 1 +Component: checks +Class: fix +Compatible: compat +Edition: cre +State: unknown +Version: 1.5.0i1 +Date: 1508241149 + +The consequence of above error was that Check_MK did not +find any ORACLE instances any more. Fixed that by using +{{nawk}} in case of SunOS. diff --git a/agents/plugins/mk_oracle b/agents/plugins/mk_oracle index dcc871d..5259008 100755 --- a/agents/plugins/mk_oracle +++ b/agents/plugins/mk_oracle @@ -742,12 +742,14 @@ sql_asm_diskgroup() function set_osenv () { ostype=$(uname -s) + AWK=$(which awk) if [ ${ostype} = 'Linux' ] ; then GREP=$(which grep) STATCX='stat -c %X' STATCY='stat -c %Y' + AWK=$(which nawk) elif [ ${ostype} = 'SunOS' ] ; then @@ -1256,7 +1258,7 @@ fi # Get list of all running databases # Do not work on ASM in this plugin. => Ignore a running ASM-Instance! -SIDS=$(UNIX95=true ps -ef | awk '{print $NF}' | ${GREP} -E '^asm_pmon_|^ora_pmon_|^xe_pmon_XE' | cut -d"_" -f3-) +SIDS=$(UNIX95=true ps -ef | ${AWK} '{print $NF}' | ${GREP} -E '^asm_pmon_|^ora_pmon_|^xe_pmon_XE' | cut -d"_" -f3-) # If we do not have found any running database instance, then either # no ORACLE is present on this system or it's just currently not running.

6 years, 11 months

Unreachable tests now test with unreachable notifications disabled and enabled

by Lars Michelsen

Module: check_mk Branch: master Commit: 0c655beecfbe367494e4d158a430f813a1afb535 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=0c655beecfbe36… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Tue Oct 17 11:30:20 2017 +0200 Unreachable tests now test with unreachable notifications disabled and enabled Change-Id: I66b25ea2a99dbfcdf6315c67859b7965ce7a93eb --- .../test_unreachable_notifications.py | 112 +++++++++++++++------ tests/testlib/__init__.py | 39 ++++++- 2 files changed, 121 insertions(+), 30 deletions(-) Diff: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commitdiff;h=0c655beecf…

6 years, 11 months

Added tests for unreachable notification logic (at default settings)

by Lars Michelsen

Module: check_mk Branch: master Commit: 04c6cb4e84219161536c6bdfade078dd780848e7 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=04c6cb4e842191… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Tue Oct 17 09:22:57 2017 +0200 Added tests for unreachable notification logic (at default settings) Change-Id: Ia33bd6c7220aef8a74b014b9f6d80ab175730e62 --- .../test_unreachable_notifications.py | 281 ++++++++++++++++----- tests/testlib/__init__.py | 42 ++- 2 files changed, 252 insertions(+), 71 deletions(-) Diff: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commitdiff;h=04c6cb4e84…

6 years, 11 months

5397 Microcore can now be configured to send UNREACHABLE notifications

by Lars Michelsen

Module: check_mk Branch: master Commit: 7826043dfc32b6c649d23cd4d70effe453271bc2 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=7826043dfc32b6… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Tue Oct 17 11:09:24 2017 +0200 5397 Microcore can now be configured to send UNREACHABLE notifications Normally you don't want to receive UNREACHABLE notifications of hosts. Configuring client/parent relationships between your hosts is a good idea to be able to detect network outages. DOWN states of hosts will then be translated into UNREACHABLE, making your view clear for the real cause (e.g. a router in DOWN state). But there may be situations where you want to be able to receive these UNREACHABLE notifications. This was already configurable when using the Nagios core for a long time. For the Microcore the suppression of UNREACHABLE notifications was hard coded. This has been changed now. The default behaviour will not be changed. New sites or updates sites will not send out UNREACHABLE notifications by default. To enable the new feature, you will first have to tell the core to send out UNREACHABLE notifications. To have a seamless transition we additionally changed these things: <ul> <li>A default rule that disables the UNREACHABLE notifications is being created for new sites in the ruleset "Notified events for hosts". Whenever you want to enable UNREACHABLE notifications, simply enable them using this rule.</li> <li>Updated sites: <ul> <li>If you have configured rules for all your hosts in the "Notified events for hosts" ruleset, the core will continue working as configured.</li> <li>Microcore: For hosts that have no matching rule in "Notified events for hosts", the Microcore will not send out UNREACHABLE notifications.</li> <li>Microcore: If you previously had configured a rule in "Notified events for hosts" which enabled the UNREACHABLE notifications, this rule had no effect. It will now be working after updating.</li> <li>Nagios: For hosts that have no matching rule in "Notified events for hosts", the Nagios core will send out UNREACHABLE notifications.</li> </ul> </li> </ul> Change-Id: Ie6dab1970f9e0dd433986e38b84a604a7a26c552 --- .werks/5397 | 44 ++++++++++++++++++++++++++++++ web/htdocs/wato.py | 23 ++++++++++++---- web/plugins/wato/check_mk_configuration.py | 10 +++++-- 3 files changed, 69 insertions(+), 8 deletions(-) diff --git a/.werks/5397 b/.werks/5397 new file mode 100644 index 0000000..631ab5a --- /dev/null +++ b/.werks/5397 @@ -0,0 +1,44 @@ +Title: Microcore can now be configured to send UNREACHABLE notifications +Level: 2 +Component: core +Compatible: compat +Edition: cee +Version: 1.5.0i1 +Date: 1508230324 +Class: feature + +Normally you don't want to receive UNREACHABLE notifications of hosts. Configuring +client/parent relationships between your hosts is a good idea to be able to +detect network outages. DOWN states of hosts will then be translated into UNREACHABLE, +making your view clear for the real cause (e.g. a router in DOWN state). + +But there may be situations where you want to be able to receive these UNREACHABLE +notifications. This was already configurable when using the Nagios core for a long +time. For the Microcore the suppression of UNREACHABLE notifications was hard coded. +This has been changed now. + +The default behaviour will not be changed. New sites or updates sites will not send +out UNREACHABLE notifications by default. To enable the new feature, you will first +have to tell the core to send out UNREACHABLE notifications. + +To have a seamless transition we additionally changed these things: + +<ul> +<li>A default rule that disables the UNREACHABLE notifications is being created + for new sites in the ruleset "Notified events for hosts". Whenever you want + to enable UNREACHABLE notifications, simply enable them using this rule.</li> +<li>Updated sites: + <ul> + <li>If you have configured rules for all your hosts in the + "Notified events for hosts" ruleset, the core will continue working + as configured.</li> + <li>Microcore: For hosts that have no matching rule in "Notified events for hosts", + the Microcore will not send out UNREACHABLE notifications.</li> + <li>Microcore: If you previously had configured a rule in "Notified events for hosts" + which enabled the UNREACHABLE notifications, this rule had no effect. It will now + be working after updating.</li> + <li>Nagios: For hosts that have no matching rule in "Notified events for hosts", + the Nagios core will send out UNREACHABLE notifications.</li> + </ul> +</li> +</ul> diff --git a/web/htdocs/wato.py b/web/htdocs/wato.py index 5bc3204..24cf9be 100644 --- a/web/htdocs/wato.py +++ b/web/htdocs/wato.py @@ -7990,9 +7990,11 @@ def event_rule_match_conditions(flavour): ( "match_host_event", ListChoice( title = _("Match host event type"), - help = _("Select the host event types and transitions this rule should handle. Note: " - "If you activate this option and do not also specify service event " - "types then this rule will never hold for service notifications!"), + help = _("Select the host event types and transitions this rule should handle. " + "Note: If you activate this option and do not also specify service event " + "types then this rule will never hold for service notifications! " + "Note: You can only match on event types <a href=\"%s\">created by the core</a>.") % \ + "wato.py?mode=edit_ruleset&varname=extra_host_conf%3Anotification_options", choices = [ ( 'rd', _("UP") + u" ➤ " + _("DOWN")), ( 'ru', _("UP") + u" ➤ " + _("UNREACHABLE")), @@ -8013,9 +8015,11 @@ def event_rule_match_conditions(flavour): ( "match_service_event", ListChoice( title = _("Match service event type"), - help = _("Select the service event types and transitions this rule should handle. Note: " - "If you activate this option and do not also specify host event " - "types then this rule will never hold for host notifications!"), + help = _("Select the service event types and transitions this rule should handle. " + "Note: If you activate this option and do not also specify host event " + "types then this rule will never hold for host notifications! " + "Note: You can only match on event types <a href=\"%s\">created by the core</a>.") % \ + "wato.py?mode=edit_ruleset&varname=extra_service_conf%3Anotification_options", choices = [ ( 'rw', _("OK") + u" ➤ " + _("WARN")), ( 'rr', _("OK") + u" ➤ " + _("OK")), @@ -14768,6 +14772,13 @@ def create_sample_config(): ], }, + # Disable unreachable notifications by default + 'extra_host_conf': { + 'notification_options': [ + ( 'd,r,f,s', [], ALL_HOSTS, {} ), + ], + }, + # Periodic service discovery 'periodic_discovery': [ ({'severity_unmonitored': 1, diff --git a/web/plugins/wato/check_mk_configuration.py b/web/plugins/wato/check_mk_configuration.py index fbdd4e0..e579999 100644 --- a/web/plugins/wato/check_mk_configuration.py +++ b/web/plugins/wato/check_mk_configuration.py @@ -1731,13 +1731,19 @@ register_rule(group, ( "f", _("Start or end of flapping state")), ( "s", _("Start or end of a scheduled downtime")), ], - default_value = [ "d", "u", "r", "f", "s" ], + default_value = [ "d", "r", "f", "s" ], ), title = _("Notified events for hosts"), help = _("This ruleset allows you to restrict notifications of host problems to certain " "states, e.g. only notify on DOWN, but not on UNREACHABLE. Please select the types " "of events that should initiate notifications. Please note that several other " - "filters must also be passed in order for notifications to finally being sent out."), + "filters must also be passed in order for notifications to finally being sent out." + " " + "Please note: There is a difference between the Microcore and Nagios when you have " + "a host that has no matching rule in this ruleset. In this case the Microcore will " + "not send out UNREACHABLE notifications while the Nagios core would send out " + "UNREACHABLE notifications. To align this behaviour, create a rule matching " + "all your hosts and configure it to either send UNREACHABLE notifications or not."), forth = lambda x: x != 'n' and x.split(",") or [], back = lambda x: ",".join(x) or "n", ),

6 years, 11 months

5344 FIX fileinfo: correct handling of file names with spaces in fileinfo.cfg

by Tom Baerwinkel

Module: check_mk Branch: master Commit: 8be294797c5792e976ce9159ae3d115b059cfac0 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=8be294797c5792… Author: Tom Baerwinkel <tb(a)mathias-kettner.de> Date: Mon Oct 16 11:54:03 2017 +0200 5344 FIX fileinfo: correct handling of file names with spaces in fileinfo.cfg Change-Id: Icdaa3afb31d14efc494e4fa99a81b13e69a98109 --- .werks/5344 | 11 +++++++++++ agents/check_mk_agent.aix | 20 ++++++++++++++------ agents/check_mk_agent.freebsd | 6 +++--- agents/check_mk_agent.linux | 8 ++++---- agents/check_mk_agent.macosx | 6 +++--- agents/check_mk_agent.openwrt | 6 +++--- agents/check_mk_agent.solaris | 13 ++++++------- 7 files changed, 44 insertions(+), 26 deletions(-) diff --git a/.werks/5344 b/.werks/5344 new file mode 100644 index 0000000..f110ce4 --- /dev/null +++ b/.werks/5344 @@ -0,0 +1,11 @@ +Title: fileinfo: correct handling of file names with spaces in fileinfo.cfg +Level: 1 +Component: checks +Class: fix +Compatible: compat +Edition: cre +State: unknown +Version: 1.5.0i1 +Date: 1508146535 + + diff --git a/agents/check_mk_agent.aix b/agents/check_mk_agent.aix index dcc7a39..7632cb5 100755 --- a/agents/check_mk_agent.aix +++ b/agents/check_mk_agent.aix @@ -287,12 +287,20 @@ fi # Fileinfo-Check: put patterns for files into /etc/check_mk/fileinfo.cfg if [ -r "$MK_CONFDIR/fileinfo.cfg" ]; then - echo '<<<fileinfo:sep(124)>>>' - date +%s - # simulate output of the Linux "stat -c "%n|%s|%Y" command - # TODO: feel free to improve this command. It opens the perl interpreter for every file found... - for pattern in $(cat "$MK_CONFDIR/fileinfo.cfg" 2>/dev/null); do if [ -f $pattern ]; then ls -1 "$pattern" ; fi; done |\ - while read file ; do /usr/bin/perl -e 'printf "$_|%s|%s\n", (stat) [7,9] for @ARGV' "$file" ; done + echo '<<<fileinfo:sep(124)>>>' + date +%s + # simulate output of the Linux "stat -c "%n|%s|%Y" command + # TODO: feel free to improve this command. It opens the perl interpreter for every file found... + while read pattern + do + if [ -f "$pattern" ]; then + ls -1 "$pattern" + fi + done < "$MK_CONFDIR/fileinfo.cfg" 2>/dev/null |\ + while read file + do + /usr/bin/perl -e 'printf "$_|%s|%s\n", (stat) [7,9] for @ARGV' "$file" + done fi # powerHA diff --git a/agents/check_mk_agent.freebsd b/agents/check_mk_agent.freebsd index 1b9a79b..3748661 100755 --- a/agents/check_mk_agent.freebsd +++ b/agents/check_mk_agent.freebsd @@ -380,14 +380,14 @@ fi if [ -r "$MK_CONFDIR/fileinfo.cfg" ] ; then echo '<<<fileinfo:sep(124)>>>' date +%s - for line in $(cat "$MK_CONFDIR/fileinfo.cfg") + while read line do - stat -f "%N|%z|%m" $line 2>/dev/null + stat -f "%N|%z|%m" "$line" 2>/dev/null if [ $? -ne 0 ]; then echo "$line|missing|$(date +%s)" fi - done + done < "$MK_CONFDIR/fileinfo.cfg" fi diff --git a/agents/check_mk_agent.linux b/agents/check_mk_agent.linux index ec7af26..8dc40c0 100755 --- a/agents/check_mk_agent.linux +++ b/agents/check_mk_agent.linux @@ -764,18 +764,18 @@ function replace_datevariable() if [ -r "$MK_CONFDIR/fileinfo.cfg" ] ; then echo '<<<fileinfo:sep(124)>>>' date +%s - for line in $(cat "$MK_CONFDIR/fileinfo.cfg") + while read line do # only work on lines containing files, skip e.g. comments and blank lines - if [ ${line:0:1} = "/" ]; then - line="$(replace_datevariable $line)" + if [ "${line:0:1}" = "/" ]; then + line=$(replace_datevariable "$line") stat -c "%n|%s|%Y" "${line}" 2>/dev/null if [ $? -ne 0 ]; then echo "$line|missing|$(date +%s)" fi fi - done + done < "$MK_CONFDIR/fileinfo.cfg" fi # Get stats about OMD monitoring cores running on this machine. diff --git a/agents/check_mk_agent.macosx b/agents/check_mk_agent.macosx index 97ec400..73b65ee 100755 --- a/agents/check_mk_agent.macosx +++ b/agents/check_mk_agent.macosx @@ -93,14 +93,14 @@ netstat -ntfinet | awk ' /^tcp/ { c[$6]++; } END { for (x in c) { print x, c[x]; if [ -r "$MK_CONFDIR/fileinfo.cfg" ] ; then echo '<<<fileinfo:sep(124)>>>' date +%s - for line in $(cat "$MK_CONFDIR/fileinfo.cfg") + while read line do - stat -f "%N|%z|%m" $line 2>/dev/null + stat -f "%N|%z|%m" "$line" 2>/dev/null if [ $? -ne 0 ]; then echo "$line|missing|$(date +%s)" fi - done + done < "$MK_CONFDIR/fileinfo.cfg" fi if type tmutil >/dev/null diff --git a/agents/check_mk_agent.openwrt b/agents/check_mk_agent.openwrt index d8ff0ec..02ae32d 100755 --- a/agents/check_mk_agent.openwrt +++ b/agents/check_mk_agent.openwrt @@ -679,14 +679,14 @@ fi if [ -r "$MK_CONFDIR/fileinfo.cfg" ] ; then echo '<<<fileinfo:sep(124)>>>' date +%s - for line in $(cat "$MK_CONFDIR/fileinfo.cfg") + while read line do - stat -c "%n|%s|%Y" $line 2>/dev/null + stat -c "%n|%s|%Y" "$line" 2>/dev/null if [ $? -ne 0 ]; then echo "$line|missing|$(date +%s)" fi - done + done < "$MK_CONFDIR/fileinfo.cfg" fi # Get stats about OMD monitoring cores running on this machine. diff --git a/agents/check_mk_agent.solaris b/agents/check_mk_agent.solaris index 05134d1..95b4dac 100755 --- a/agents/check_mk_agent.solaris +++ b/agents/check_mk_agent.solaris @@ -369,15 +369,14 @@ then echo '<<<fileinfo:sep(124)>>>' /usr/bin/perl -e 'print time."\n"' # note: patterns in the fileinfo.cfg are resolved inside this loop - for file in $(cat "$MK_CONFDIR/fileinfo.cfg") + while read file do # only work on lines containing files, skip e.g. comments and blank lines - if [ ${file:0:1} = "/" ]; then - file="$(replace_datevariable $file)" - ls $file > /dev/null 2>&1 - + if [ "${file:0:1}" = "/" ]; then + file=$(replace_datevariable "$file") + ls "$file" > /dev/null 2>&1 if [ $? -eq 0 ]; then - echo $file | \ + echo "$file" | \ /usr/bin/perl -e ' while(my $file = <>) { $file =~ s/\n$//; @@ -388,7 +387,7 @@ then echo "$file|missing|$(/usr/bin/perl -e 'print time')" fi fi - done + done < "$MK_CONFDIR/fileinfo.cfg" fi # Libelle Business Shadow

6 years, 11 months

5396 FIX Nagios: Log external commands to nagios.log by default

by Lars Michelsen

Module: check_mk Branch: master Commit: 8f4911d356e42e965b01533be5ca8be345454a3c URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=8f4911d356e42e… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Fri Oct 13 11:37:16 2017 +0200 5396 FIX Nagios: Log external commands to nagios.log by default Nagios is now logging external commands to the core log (nagios.log) in the default configuration. This makes Nagios behave like the Microcore in this case. Change-Id: Ie18d93111a62102753fa0a32547cbfdde1c416de --- .werks/5396 | 12 ++++++++++++ omd/packages/nagios/skel/etc/nagios/nagios.d/logging.cfg | 2 +- 2 files changed, 13 insertions(+), 1 deletion(-) diff --git a/.werks/5396 b/.werks/5396 new file mode 100644 index 0000000..71c5cab --- /dev/null +++ b/.werks/5396 @@ -0,0 +1,12 @@ +Title: Nagios: Log external commands to nagios.log by default +Level: 1 +Component: core +Compatible: compat +Edition: cre +Version: 1.5.0i1 +Date: 1507887368 +Class: fix + +Nagios is now logging external commands to the core log (nagios.log) +in the default configuration. This makes Nagios behave like the Microcore +in this case. diff --git a/omd/packages/nagios/skel/etc/nagios/nagios.d/logging.cfg b/omd/packages/nagios/skel/etc/nagios/nagios.d/logging.cfg index 4e98868..539f87f 100644 --- a/omd/packages/nagios/skel/etc/nagios/nagios.d/logging.cfg +++ b/omd/packages/nagios/skel/etc/nagios/nagios.d/logging.cfg @@ -65,7 +65,7 @@ log_initial_states=1 # checks - see the option below for controlling whether or not # passive checks are logged. -log_external_commands=0 +log_external_commands=1 # PASSIVE CHECKS LOGGING OPTION # If you don't want Nagios to log passive host and service checks, set

6 years, 11 months

5395 FIX Nagios: Enable passive host check translation by default

by Lars Michelsen

Module: check_mk Branch: master Commit: a78a1fd6fb0be5158e856157e08b71775d218512 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=a78a1fd6fb0be5… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Fri Oct 13 11:35:34 2017 +0200 5395 FIX Nagios: Enable passive host check translation by default The monitoring cores Nagios and Microcore where treating "fake check results" that can be sent from the GUI differently. These fake check results are in fact passive host checks. From these passive host checks Nagios simply takes over the reported state and uses it as new host state, ignoring the parent logic. The microcore takes the parent logic into account and may change a DOWN check result to UNREACHABLE when all parents are down. We have now changed the configuration option <tt>translate_passive_host_checks</tt> of Nagios to be enabled by default to make Nagios also check this logic. Change-Id: Ica73530773f18c2720088a99aa83f093535ffb5b --- .werks/5395 | 19 +++ .../nagios/skel/etc/nagios/nagios.d/freshness.cfg | 2 +- .../test_unreachable_notifications.py | 144 +++++++++++++++++++++ tests/testlib/__init__.py | 24 ++++ tests/web/test_webapi.py | 2 +- 5 files changed, 189 insertions(+), 2 deletions(-) diff --git a/.werks/5395 b/.werks/5395 new file mode 100644 index 0000000..be92538 --- /dev/null +++ b/.werks/5395 @@ -0,0 +1,19 @@ +Title: Nagios: Enable passive host check translation by default +Level: 1 +Component: core +Compatible: compat +Edition: cre +Version: 1.5.0i1 +Date: 1507886901 +Class: fix + +The monitoring cores Nagios and Microcore where treating "fake check results" +that can be sent from the GUI differently. + +These fake check results are in fact passive host checks. From these passive host +checks Nagios simply takes over the reported state and uses it as new host state, +ignoring the parent logic. The microcore takes the parent logic into account and +may change a DOWN check result to UNREACHABLE when all parents are down. + +We have now changed the configuration option <tt>translate_passive_host_checks</tt> +of Nagios to be enabled by default to make Nagios also check this logic. diff --git a/omd/packages/nagios/skel/etc/nagios/nagios.d/freshness.cfg b/omd/packages/nagios/skel/etc/nagios/nagios.d/freshness.cfg index c5fa11a..3360260 100644 --- a/omd/packages/nagios/skel/etc/nagios/nagios.d/freshness.cfg +++ b/omd/packages/nagios/skel/etc/nagios/nagios.d/freshness.cfg @@ -12,7 +12,7 @@ # passively into the correct state from the view of this server. # Values: 1 = perform translation, 0 = do not translate (default) -translate_passive_host_checks=0 +translate_passive_host_checks=1 diff --git a/tests/notifications/test_unreachable_notifications.py b/tests/notifications/test_unreachable_notifications.py new file mode 100644 index 0000000..67e62dc --- /dev/null +++ b/tests/notifications/test_unreachable_notifications.py @@ -0,0 +1,144 @@ +#!/usr/bin/env python +# encoding: utf-8 + +import pytest +import time +from testlib import web + +(a)pytest.fixture(scope="module") +def test_cfg(web, site): + try: + print "Applying test config" + + web.add_host("notify-test-parent", attributes={ + "ipaddress": "127.0.0.1", + }) + + web.add_host("notify-test-child", attributes={ + "ipaddress": "127.0.0.1", + "parents": [ "notify-test-parent" ], + }) + + web.activate_changes() + + site.live.command("[%d] DISABLE_HOST_CHECK;notify-test-parent" % time.time()) + site.live.command("[%d] DISABLE_HOST_CHECK;notify-test-child" % time.time()) + + yield None + finally: + # + # Cleanup code + # + print "Cleaning up default config" + + web.delete_host("notify-test-child") + web.delete_host("notify-test-parent") + + +def set_initial_state(site): + # Before each test: Set to initial state: Both UP + site.send_host_check_result("notify-test-child", 0, "UP") + site.send_host_check_result("notify-test-parent", 0, "UP") + + # Before each test: Clear logs + site.live.command("[%d] ROTATE_LOGFILE" % time.time()) + time.sleep(1) # TODO: Add check for rotation + + +def open_history_log(core): + if core == "cmc": + return open("var/check_mk/core/history") + elif core == "nagios": + return open("var/nagios/nagios.log") + else: + raise NotImplementedError() + + +STATE_UP = 0 +STATE_DOWN = 1 +STATE_UNREACHABLE = 2 + +(a)pytest.mark.parametrize(("core"), [ "nagios", "cmc" ]) +def test_unreachable_child_down_before_parent_down(test_cfg, site, core): + site.set_core(core) + set_initial_state(site) + + # TODO: + # - Set child down, expect DOWN notification + site.send_host_check_result("notify-test-child", STATE_DOWN, "DOWN") + + assert "HOST ALERT: notify-test-child;DOWN;HARD;1;DOWN" in open_history_log(core).read() + + # - Set parent down, expect DOWN notification for parent and UNREACHABLE notification for child + site.send_host_check_result("notify-test-parent", STATE_DOWN, "DOWN") + + # Difference beween nagios/cmc: when sending DOWN via PROCESS_HOST_CHECK_RESULT + # the nagios core needs another child down check result to report it as unreachable. + if core == "nagios": + site.send_host_check_result("notify-test-child", STATE_DOWN, "DOWN", expected_state=STATE_UNREACHABLE) + + history_log = open_history_log(core).read() + + assert "HOST ALERT: notify-test-parent;DOWN;HARD;1;DOWN" in history_log + assert "HOST ALERT: notify-test-child;UNREACHABLE;HARD;1;" in history_log + + if core == "cmc": + assert "HOST NOTIFICATION: check-mk-notify;notify-test-parent;DOWN;check-mk-notify;" in history_log + assert "HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;" in history_log + else: + # TODO: Nagios does not log the entries checked above for cmc. This may be a problem e.g. for availability. + pass + + +(a)pytest.mark.parametrize(("core"), [ "nagios", "cmc" ]) +def test_unreachable_child_after_parent_is_down(test_cfg, site, core): + site.set_core(core) + set_initial_state(site) + + + # TODO: + # - Set parent down, expect DOWN notification + site.send_host_check_result("notify-test-parent", STATE_DOWN, "DOWN") + + assert "HOST ALERT: notify-test-parent;DOWN;HARD;1;DOWN" in open_history_log(core).read() + + # - set child down, expect UNREACHABLE notification + assert site.get_host_state("notify-test-child") == STATE_UP + site.send_host_check_result("notify-test-ychild", STATE_DOWN, "DOWN") + assert site.get_host_state("notify-test-child") == STATE_UNREACHABLE + + history_log = open_history_log(core).read() + + assert "HOST ALERT: notify-test-child;DOWN;HARD;1;DOWN" in history_log + + #if core == "cmc": + # assert "HOST ALERT: notify-test-child;UNREACHABLE;HARD;1;child becomes unreachable due to state change of parent host" in history_log + + # assert "HOST NOTIFICATION: check-mk-notify;notify-test-parent;DOWN;check-mk-notify;" in history_log + # assert "HOST NOTIFICATION: check-mk-notify;notify-test-child;UNREACHABLE;check-mk-notify;" in history_log + #else: + # # TODO: Nagios does not log the entries checked above for cmc. This may be a problem e.g. for availability. + # pass + + # TODO: + # - Set parent down, expect DOWN notification + # - Set child up, expect no notification + + # TODO: + # - Set parent down, expect DOWN notification + # - set child down, expect UNREACHABLE notification + # - set child up, expect UP notification + # - set child down, expect UNREACHABLE notification + + # TODO: + # - Set parent down, expect DOWN notification + # - set child down, expect UNREACHABLE notification + # - Set parent up, expect UP notification and child DOWN notification + # - Set parent down, expect DOWN notification and UNREACHABLE child notification + + +#(a)pytest.mark.parametrize(("core"), [ "nagios", "cmc" ]) +#def test_unreachable_disabled_by_default(test_cfg, site, core): +# # TODO: Set child down, set parent down +# # TODO: Check log that no UNREACHABLE notification has been created +# pass diff --git a/tests/testlib/__init__.py b/tests/testlib/__init__.py index 248dc6d..74a9886 100644 --- a/tests/testlib/__init__.py +++ b/tests/testlib/__init__.py @@ -239,6 +239,26 @@ class Site(object): return live + def send_host_check_result(self, hostname, state, output, expected_state=None): + if expected_state is None: + expected_state = state + + self.live.command("[%d] PROCESS_HOST_CHECK_RESULT;%s;%d;%s" % (time.time(), hostname, state, output)) + + # Wait for processed command + timeout = 10 + host_state = self.get_host_state(hostname) + while timeout and not host_state == expected_state: + timeout -= 1 + time.sleep(1) + host_state = self.get_host_state(hostname) + assert host_state == expected_state, "Expected %d state, got %d state" % (expected_state, host_state) + + + def get_host_state(self, hostname): + return self.live.query_value("GET hosts\nColumns: state\nFilter: host_name = %s" % hostname) + + def _is_running_as_site_user(self): return pwd.getpwuid(os.getuid()).pw_name == self.id @@ -579,6 +599,10 @@ class Site(object): print "Started site" + def set_core(self, core): + self.set_config("CORE", core, with_restart=True) + + def get_config(self, key): p = self.execute(["omd", "config", "show", key], stdout=subprocess.PIPE, stderr=subprocess.PIPE) stdout, stderr = p.communicate() diff --git a/tests/web/test_webapi.py b/tests/web/test_webapi.py index 9b5540d..b1fcc27 100644 --- a/tests/web/test_webapi.py +++ b/tests/web/test_webapi.py @@ -358,7 +358,7 @@ def test_get_graph(web, site): web.activate_changes() # Issue a reschedule - site.live.command("SCHEDULE_FORCED_SERVICE_CHECK;test-host-get-graph;Check_MK;%d" % int(time.time())) + site.live.command("[%d] SCHEDULE_FORCED_SERVICE_CHECK;test-host-get-graph;Check_MK;%d" % (int(time.time), int(time.time()))) # Wait for RRD file creation # Isn't this a bug that the graph is not instantly available?

6 years, 11 months

5341 agent_vsphere: change of the option "Display VM power state on"

by Tom Baerwinkel

Module: check_mk Branch: master Commit: be06290af8c3292aafa17972588bb6cf682a4660 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=be06290af8c329… Author: Tom Baerwinkel <tb(a)mathias-kettner.de> Date: Thu Oct 12 11:37:22 2017 +0200 5341 agent_vsphere: change of the option "Display VM power state on" The option "Display VM power state on" of the WATO rule now displays the power state of VMs additionally on either the ESX host or the VM. The option is therefore renamed to "Display power state additionally on". This will result in services for both the queried system and the ESX host / VM. By disabling the unwanted services it is then possible to configure where the corresponding services are displayed. Because the data is now still available on the queried system the service "Object count" will now display the correct count when the option is used. Change-Id: I89172eaa9384a7ef0436a1084ce812ae9cdd2d07 --- .werks/5341 | 18 ++++++++++++++++++ agents/special/agent_vsphere | 6 ++---- web/plugins/wato/datasource_programs.py | 8 +++++++- 3 files changed, 27 insertions(+), 5 deletions(-) diff --git a/.werks/5341 b/.werks/5341 new file mode 100644 index 0000000..d4e6821 --- /dev/null +++ b/.werks/5341 @@ -0,0 +1,18 @@ +Title: agent_vsphere: change of the option "Display VM power state on" +Level: 1 +Component: checks +Compatible: compat +Edition: cre +Version: 1.5.0i1 +Date: 1507799865 +Class: feature + +The option "Display VM power state on" of the WATO rule now displays +the power state of VMs additionally on either the ESX host or +the VM. The option is therefore renamed to "Display power state +additionally on". This will result in services for both the +queried system and the ESX host / VM. By disabling the unwanted +services it is then possible to configure where the corresponding +services are displayed. Because the data is now still available on +the queried system the service "Object count" will now display the +correct count when the option is used. diff --git a/agents/special/agent_vsphere b/agents/special/agent_vsphere index cebc547..e272957 100755 --- a/agents/special/agent_vsphere +++ b/agents/special/agent_vsphere @@ -1662,8 +1662,7 @@ if not error: host_info = "hostsystem\t%s\t\t%s" % (converted_hostname, hostsystems_properties[hostname]["runtime.powerState"][0]) if opt_host_pwr_display == "esxhost" and not opt_direct: host_piggy_data.setdefault(converted_hostname, []).append(host_info) - else: - output(host_info) + output(host_info) if "virtualmachine" in query_objects: for key in vms.keys(): @@ -1676,8 +1675,7 @@ if not error: vm_piggy_data.setdefault(vm_name, []).append(vm_info) elif opt_vm_pwr_display == "esxhost" and not opt_direct: host_piggy_data.setdefault(running_on, []).append(vm_info) - else: - output(vm_info) + output(vm_info) if "virtualmachine" in query_objects and "hostsystem" in query_objects and opt_host_pwr_display == "vm": def find_host(hostname): diff --git a/web/plugins/wato/datasource_programs.py b/web/plugins/wato/datasource_programs.py index 7dbdbda..da9539c 100644 --- a/web/plugins/wato/datasource_programs.py +++ b/web/plugins/wato/datasource_programs.py @@ -158,7 +158,13 @@ register_rule(group, ), ( "vm_pwr_display", DropdownChoice( - title = _("Display VM power state on"), + title = _("Display VM power state additionally on"), + help = _("The power state can be displayed additionally either " + "on the ESX host or the VM. This will result in services " + "for both the queried system and the ESX host / VM. " + "By disabling the unwanted services it is then possible " + "to configure where the services are displayed." + ), choices = [ ( None, _("The queried ESX system (vCenter / Host)") ), ( "esxhost", _("The ESX Host") ),

6 years, 11 months

5226 netstat, win_netstat: Checks now include performance graphs for number of connections

by Konstantin Büttner

Module: check_mk Branch: master Commit: b0f960688490c3da722414c9da294bc632449547 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=b0f960688490c3… Author: Konstantin Büttner <kb(a)mathias-kettner.de> Date: Fri Oct 13 13:42:24 2017 +0200 5226 netstat, win_netstat: Checks now include performance graphs for number of connections Change-Id: If0187520250194083828fe0278dc12a7ae87defd --- .werks/5226 | 10 ++++++++++ checks/netstat | 1 + checks/netstat.include | 5 ++++- checks/win_netstat | 1 + 4 files changed, 16 insertions(+), 1 deletion(-) diff --git a/.werks/5226 b/.werks/5226 new file mode 100644 index 0000000..94ee88e --- /dev/null +++ b/.werks/5226 @@ -0,0 +1,10 @@ +Title: netstat, win_netstat: Checks now include performance graphs for number of connections +Level: 1 +Component: checks +Compatible: compat +Edition: cre +Version: 1.5.0i1 +Date: 1507894897 +Class: feature + + diff --git a/checks/netstat b/checks/netstat index 6176f3f..a7d6975 100644 --- a/checks/netstat +++ b/checks/netstat @@ -84,5 +84,6 @@ check_info["netstat"] = { 'service_description' : "TCP Connection %s", 'group' : "tcp_connections", 'includes' : [ "netstat.include" ], + 'has_perfdata' : True, } diff --git a/checks/netstat.include b/checks/netstat.include index a57cdad..1b95bdc 100644 --- a/checks/netstat.include +++ b/checks/netstat.include @@ -69,6 +69,7 @@ def check_netstat_generic(item, params, connections): if params.get("max_states"): max_warn, max_crit = params["max_states"] + perfdata = [ ("connections", found, max_warn, max_crit) ] if found >= max_crit: max_state = 2 elif found >= max_warn: @@ -78,6 +79,8 @@ def check_netstat_generic(item, params, connections): if max_state: states.append(max_state) infotext += " (warn/crit at %d/%d)" % (max_warn, max_crit) + else: + perfdata = [ ("connections", found) ] - yield max(states), infotext + yield max(states), infotext, perfdata diff --git a/checks/win_netstat b/checks/win_netstat index 3853095..f7978fd 100644 --- a/checks/win_netstat +++ b/checks/win_netstat @@ -87,4 +87,5 @@ check_info["win_netstat"] = { 'service_description' : "TCP Connection %s", 'group' : "tcp_connections", 'includes' : [ "netstat.include" ], + 'has_perfdata' : True, }

6 years, 11 months

← Newer
1
...
14
15
16
17
18
19
20
...
29
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Checkmk git commits October 2017