Module: check_mk
Branch: master
Commit: b0ae55862e694d25366bded09f1f6306b8dde5c7
URL:
http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=b0ae55862e694d…
Author: Lars Michelsen <lm(a)mathias-kettner.de>
Date: Wed Sep 6 14:49:00 2017 +0200
5185 FIX Check_MK HW/SW Inventory: The service was executed too often by default
Issue: The HW/SW inventory may produce a lot of load because of wrong scheduling.
You may be affected by this bug in case you created your (master) site with 1.4.0b1
or newer. If you updated your site from older versions you are likely not affected.
Slave sites that get their configuration from a master that is not affected will also
not be affected, regardless of the version they were created with.
When using the hardware software inventory the regular check "Check_MK HW/SW
Inventory"
on each host is executed too often by default. The expected behaviour is that the regular
check interval of this service is 1 day. The bug we fixed now made the check being
executed
once a minute. In case you have the HW/SW inventory enabled for a lot of hosts, this
could
lead to significant load on your Check_MK server.
Whether or not you have this issue can be checked on the "service detail" page
of the
"Check_MK HW/SW Inventory" service. Have a look at the "Service
normal/retry check interval"
column where you should see 86400s as first value (-> 1 day). If you see a value of
60s, you
should fix the issue as follows.
<ul>
<li>Navigate to "WATO > Host & Service parameters > Monitoring
configuration > Normal check
interval for service checks".</li>
<li>Create a new rule with the following settings:</li>
<li>Description: Restrict HW/SW-Inventory to once a day</li>
<li>Value: 1 day</li>
<li>Services: <tt>Check_MK HW/SW Inventory$</tt></li>
<li>Activate the changes via WATO</li>
<li>
After this you should check the service detail page again and see the 86400s interval.
Even when you update your existing site to 1.4.0p11 or newer, you will have to apply the
fix
manually to existing installations when your installation is affected.
Sites created with 1.4.0p11 or newer won't be affected.
Change-Id: Iaf50c46d1c92a73e0051e35b1bba54508eab51f6
---
.werks/5185 | 43 +++++++++++++++++++++++++++++++++++++++++++
web/htdocs/wato.py | 8 +++++---
2 files changed, 48 insertions(+), 3 deletions(-)
diff --git a/.werks/5185 b/.werks/5185
new file mode 100644
index 0000000..24a4591
--- /dev/null
+++ b/.werks/5185
@@ -0,0 +1,43 @@
+Title: Check_MK HW/SW Inventory: The service was executed too often by default
+Level: 2
+Component: inv
+Class: fix
+Compatible: incomp
+Edition: cre
+State: unknown
+Version: 1.5.0i1
+Date: 1504701323
+
+Issue: The HW/SW inventory may produce a lot of load because of wrong scheduling.
+
+You may be affected by this bug in case you created your (master) site with 1.4.0b1
+or newer. If you updated your site from older versions you are likely not affected.
+Slave sites that get their configuration from a master that is not affected will also
+not be affected, regardless of the version they were created with.
+
+When using the hardware software inventory the regular check "Check_MK HW/SW
Inventory"
+on each host is executed too often by default. The expected behaviour is that the
regular
+check interval of this service is 1 day. The bug we fixed now made the check being
executed
+once a minute. In case you have the HW/SW inventory enabled for a lot of hosts, this
could
+lead to significant load on your Check_MK server.
+
+Whether or not you have this issue can be checked on the "service detail" page
of the
+"Check_MK HW/SW Inventory" service. Have a look at the "Service
normal/retry check interval"
+column where you should see 86400s as first value (-> 1 day). If you see a value of
60s, you
+should fix the issue as follows.
+
+<ul>
+<li>Navigate to "WATO > Host & Service parameters > Monitoring
configuration > Normal check
+interval for service checks".</li>
+<li>Create a new rule with the following settings:</li>
+<li>Description: Restrict HW/SW-Inventory to once a day</li>
+<li>Value: 1 day</li>
+<li>Services: <tt>Check_MK HW/SW Inventory$</tt></li>
+<li>Activate the changes via WATO</li>
+<li>
+
+After this you should check the service detail page again and see the 86400s interval.
+
+Even when you update your existing site to 1.4.0p11 or newer, you will have to apply the
fix
+manually to existing installations when your installation is affected.
+Sites created with 1.4.0p11 or newer won't be affected.
diff --git a/web/htdocs/wato.py b/web/htdocs/wato.py
index 08ffc9c..7501ad1 100644
--- a/web/htdocs/wato.py
+++ b/web/htdocs/wato.py
@@ -14765,9 +14765,11 @@ def create_sample_config():
],
# Interval for HW/SW-Inventory check
- 'extra_service_conf:check_interval': [
- ( 1440, [], ALL_HOSTS, [ "Check_MK HW/SW Inventory$" ],
{'description': u'Restrict HW/SW-Inventory to once a day'} ),
- ],
+ 'extra_service_conf': {
+ 'check_interval': [
+ ( 1440, [], ALL_HOSTS, [ "Check_MK HW/SW Inventory$" ],
{'description': u'Restrict HW/SW-Inventory to once a day'} ),
+ ],
+ },
# Periodic service discovery
'periodic_discovery': [