ID: 1406
Title: Assume PEND in count_ok aggregations if all nodes are PEND
Component: BI
Level: 2
Class: New Feature
Version: 1.2.5i6
If you are using the BI aggregation function <tt>count_ok</tt> then nodes
with the state <tt>PEND</tt> (which essentially means unmonitored or not
present) are handled differently from now.
If <b>all</b> nodes are pending then the total state will also be pending.
If just some of the nodes are pending then they will not taken into account.
This is especially imported when doing availability computations on
BI aggregates.
ID: 1405
Title: Checkbox for settings downtimes on the hosts of the selected services
Component: Multisite
Level: 2
Class: New Feature
Version: 1.2.5i6
In the Multisite command box for setting downtimes on services there is now
a new checkbox called <i>Schedule downtimes on the affected hosts instead
of their services</i>. When you check this then the scheduled downtimes
are set on the hosts of the selected services rather then on the services
themselves. Note: When a host is in scheduled downtime then all its services
are considered to be in a scheduled downtime, as well. There is no need to
set a downtime on these.
ID: 1117
Title: agent_vsphere: now able to query license information from esx system
Component: Checks & Agents
Level: 2
Class: New Feature
Version: 1.2.5i6
The agent_vsphere is now able to get the license usage from the LicenseManager instance.
Upon inventory each license will lead to an own check name esx_licenses.
The check parameters (absolute used, percentage used, all used) can be configured via the WATO rule
<tt>Number of used VMware licenses</tt>
ID: 1403
Title: kernel.util: allow levels for the total CPU utilization
Component: Checks & Agents
Level: 2
Class: New Feature
Version: 1.2.5i6
The CPU utilization check for Linux only allowed to set levels for the disk
wait percentage. The parameters have been converted to a dictionary and
now allow to set levels on the total CPU utilization as well.
Please note: if you have defined the variable {kernel_util_default_levels} in
your {main.mk}, then you need to convert this to a dictionary - otherwise it
will be ignored. Instead of
F+:main.mk
kernel_util_default_levels = (10, 20)
F-:
... please now write...:
F+:main.mk
kernel_util_default_levels = { "iowait" : (10, 20) }
F-:
Users of WATO are not affected by this and do not need to change anything.
ID: 1402
Title: Fixed cased where counting events did not reach required count
Component: Event Console
Level: 2
Class: Bug Fix
Version: 1.2.5i6
This bug was found and fixed by Andreas Döhler!
ID: 1401
Title: Display options in views are now again persistent
Component: Multisite
Level: 2
Class: Bug Fix
Version: 1.2.5i6
When you change a "Display Option" in a view - such as the <i>Time stamp format</i>, then
this option was only valid whilst you were visiting that view. This was due to a bug, which
was introduced in Januar 2014. This has been fixed now. If you are visiting a view again
after changing display options, then these will be persisted forever. Note: the persistance
is on a per-user-base and per-view-base.
ID: 1400
Title: Added example notification script for Pushover to doc/treasures/notifications
Component: Notifications
Level: 2
Class: New Feature
Version: 1.2.5i6
This plugin can send notifications via pushover.net. You need to edit your keys
directly in the script. There is not yet a WATO integration.
ID: 1111
Title: vCenter monitoring: greatly improved performance (at least 40 times faster)
Component: Checks & Agents
Level: 3
Class: New Feature
Version: 1.2.5i6
The pysphere API is no longer used when monitoring ESX vCenter systems.
We are now using much simpler tools to evaluate the vCenter response instead,
thus saving quite a lot of cpu time. Depending on the number of ESX Hosts and
virtualmachines you will see a performance increase by a factor of 40-200.
<br><br>
Note:<br>
If the "Compatibility mode" is set in WATO the old "pysphere" version is still being used.
This option remains for older (<5.0) ESX versions.
ID: 1399
Title: Fix left-over tac processes when showing Event Console history
Component: Event Console
Level: 2
Class: Bug Fix
Version: 1.2.5i6
When you show the Event Console history and run into a soft or hard
query limit (of usually 1,000 or 5,000 entries), then every time one
<tt>tac</tt> process was left over. This has been fixed.
ID: 1398
Title: Allow to spread times of next check when rescheduling
Component: Multisite
Level: 2
Class: New Feature
Version: 1.2.5i6
When you do the action <i>Reschedule active checks</i> for more than one
active checks then you now can enter a number of minutes. All checks will
then be distributed evenly over that time.
This feature is interesting for checks with larger intervals. Most prominent
example is the <tt>Check_MK Inventory</tt> check, which usually is done every
2, 4 or 24 hours. By spreading the executing of these checks you avoid large
peaks in CPU load.