ID: 6014
Title: Fixed handling of expected regular messages
Component: Event Console
Level: 2
Class: Bug fix
Version: 1.6.0i1
If expected regular messages did not arrive at the event console and the
resulting new event was not merged with a previous one, the EC could go into
an endless loop, logging Python exceptions all the time. This has been
fixed.
ID: 6191
Title: Configure additional Windows eventlogs with keyword 'logfile'
Component: Checks & agents
Level: 2
Class: Bug fix
Version: 1.6.0i1
Werk #3106 (Check_MK 1.4.0i1) introduced the possibility of configuring
additional eventlogs with the keyword 'logname'. However, this was not
supported by the Agent Bakery that only knew the keyword 'logfile'.
Now, additional, non-standard eventlogs that are not found in Windows
registry, can be configured for monitoring with the ordinary keyword
'logfile', e.g.
"logfile Microsoft-Windows-GroupPolicy/Operational = warn"
so that they can also be configured normally through the Agent Bakery.
Note that non-standard eventlogs also require the option vista_api to
be set to 'yes.'
Using the keyword 'logname' is strongly discouraged and that keyword
should be considered deprecated. It will remain functional only for
compatibility with old, manually configured check_mk.ini files.
ID: 5823
Title: New Feature: SLA reporting
Component: Multisite
Level: 3
Class: New feature
Version: 1.6.0i1
SLA reports are a continuation of the already existing availability reports.
The data from the availability feature can be additionally modified, before it is evaluated by the SLA requirements.
Currently these two SLA requirements are implemented
<ul>
<li>State Percentage: The service state OK/WARN/CRIT/UNKNOWN is lower/higher than x percent</li>
<li>Outage count: Maximum number of outages with a length of x minutes</li>
</ul>
Note: An upcoming SLA requirement will be "Minimum time between outages"
Before the availability data gets forwarded to the computation plugins (which check the actual requirements) it is
<li>further shortened by applying timeperiods, for example "Only check weekdays"</li>
<li>cut into mulitple timeframes (daily/weekly/monthly), specified by period in the SLA definition</li>
</ul>
Each timeframe is computed separately, so you will get several SLA results, if the timerange you are interested in spans multiple SLA periods.
Configuration
First of all you need to configure a SLA definition. Just like views and reports, SLA definitions are configured per user,
so you can setup a SLA and publish it to other users.
SLA definitions are not inherently bound to services.
A SLA definition can be assigned to a service, via the WATO rule <tt>Assign SLA to service</tt>.
For views, the multisite GUI currently offers two different painters to display SLA information
<ul>
<li>SLA - Service specific: This painter renders SLA definitions, which were assigned by WATO rule. Therefore, one table
column in a view can display multiple SLA definition types. The timerange, however, is fixed for all SLAs and can be configured in the painter</li>
<li>SLA - Column specific: This painter renders the same SLA definition for all services in the table</li>
</ul>
There are several configuration options for both painters.
<ul>
<li>SLA timerange: The timerange specifies the start and end date you are interested in. A SLA definition only has a
recurring period (e.g. weekly). So if you have a SLA definition with a monthly period and set the timerange from
01.01.2018 to 31.03.2018, you will get three separate SLA results, one for each month</li>
<li>Layout options: With these options you can configure the content of the painter. You may also choose only to display
the SLA name, which means no actual SLA computation, hereby saving performance</li>
<li>Plugin display options</li>Allows you to configure the precision of percentage (float) values and additional layout options</li>
</ul>
There is also a SLA details page available. This page offers detailed information regarding the computation.
<ul>
<li>Availability rawdata used in computation</li>
<li>Aggregated results (the actual SLA outcome) over all computation plugins</li>
<li>Subresults for each used computation plugin</li>
<li>SLA specification: Includes all settings which were used to create this report</li>
</ul>
Please note that this feature is quite new and still under heavy development.
The goal in version 1.5 is to get some hands on experience and add some minor improvements and bugfixes over time.
Exporting the SLA data into PDF reports is currently only supported within views.
Later on, the SLA details page will also receive a "Export to PDF" button.
ID: 5816
Title: Check parse_function is no longer called multiple times if there are several subchecks for the same section
Component: Checks & agents
Level: 2
Class: Bug fix
Version: 1.6.0i1
ID: 5814
Title: Fixed missing clustered snmp services on cluster hosts
Component: Checks & agents
Level: 2
Class: Bug fix
Version: 1.6.0i1
Clustered snmp services did not receive from their nodes.
The data required by the cluster host was either missing or an error message was displayed, stating:
"The check type filter function has not been set".
ID: 6107
Title: Agent bakery signing key passphrases could be visible in access logs
Component: WATO
Level: 2
Class: Security fix
Version: 1.6.0i1
When you are using the agent bakery for creating and distributing your
monitoring agents it is likely that this change is relevant for you.
In some parts of the GUI, after entering the passphrase of the agent signing
keys, it could happen that the signing key passphrase you enter is written to
the apache access log of your Check_MK server. As a result it may be visible in
clear text to local system users (e.g. users with access to the command line)
which scan the logs for it.
This affects the access log of the system apache (normally located at
/var/log/apache2) and the access log (/omd/sites/[site]/var/log/apache/*) of
the sites apache (master site in distributed setups).
You may want to scan the log files for the string "key_p_passphrase" to check
whether or not you are affected. It can be done e.g. with:
zgrep key_p_passphrase /var/log/apache2/access*
/omd/sites/*/var/log/apache/access*
In case you find something, you should clean it up. Delete the file or remove
the secrets from that file.
Even when it seems unlikely that your key has been compromised, it is
recommended to stop using this signing key. Create a new key and proceed with
this one.
ID: 6170
Title: Raw Edition: Fixed monitoring using special agents
Component: Core & setup
Level: 2
Class: Bug fix
Version: 1.6.0i1
When trying to monitor hosts using special agents as data sources with nagios, this
failed with an error like this:
"UNKN - [special_vsphere] 'vsphere'UNKN, Got no information from host, execution time 0.0 sec"
ID: 6154
Title: Fixed different Check_MK calls (parent scan, baking, rrd conversion) in some cases
Component: Core & setup
Level: 2
Class: Bug fix
Version: 1.6.0i1
When using some check specific configuration variables (like e.g. fileinfo_groups), the
mentioned Check_MK calls could not be used. An error like this was shown:
Cannot read in configuration file /omd/sites/beta/etc/check_mk/conf.d/wato/rules.mk: name 'fileinfo_groups' is not defined
ID: 6153
Title: Fixed broken notifications and alert handling (Regression in 1.5.0b5)
Component: Notifications
Level: 2
Class: Bug fix
Version: 1.6.0i1
The notifications and alert handling were not working correctly at all. Better
use the 1.5.0b5 only in test environments where you don't need these features.