ID: 7384
Title: check_http: Improve WATO rule
Component: Checks & agents
Level: 1
Class: New feature
Version: 1.6.0i1
The WATO rule to configure the active check 'check_http' now provides a more
user friendly way to configure proxy settings.
ID: 7250
Title: Check_MK Discovery: Clusters were unable to automatically discover new services
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 1.6.0i1
The <tt>Check_MK Discovery</tt> check for a cluster was able to identify new/vanished services.
However, the automatic discovery mechanism in the background failed to integrate them into the configuration.
This has been fixed. There is still some constraint which makes this werk incompatible.
The <tt>Periodic service discovery</tt> rule allows you to configure <tt>Automatcally activate changes</tt> for the specified host.
If you want automatic activation of new services for cluster hosts, you have to configured a rule on the nodes instead of the cluster hosts.
So any <tt>Periodic service discovery</tt> rules of clusters should be duplicated onto the nodes.
This requirement is currently necessary because of the internal configuration.
Clusters never have an explicit configuration file, the actual clusters configuration is compute on demand.
ID: 7146
Title: mk_docker.py: Optionally use container names as identifier
Component: Checks & agents
Level: 1
Class: New feature
Version: 1.6.0i1
You can now configure the plugin mk_docker.py to use the containers
name or long ID as host name. You can do this either using the
agent bakery rule "Docker node and containers", or via configuration
file.
ID: 7222
Title: check_http: Mixup of ports when configuring certificate check via proxy
Component: WATO
Level: 1
Class: Bug fix
Version: 1.6.0i1
If users had the active check "Check HTTP" configured to check a certificate
via a proxy, the option "TCP Port" had wrongly been used as the proxys port
instead of the certificate servers port.
The port defined in the option "TCP Port" is now applied to the certificate
server, regardless of whether a proxy is used.
If you relied on the wrong behaviour, remove the option "TCP Port" and include
the port for the proxy in the proxy settings.
ID: 7223
Title: db2_tablespaces: Occasionally wrong CRIT state
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 1.6.0i1
In some cases the state of the check went to CRIT, although the
reported values where fine.
ID: 7249
Title: Check_Discovery: Automatic discovery of hosts now correctly ignores unreachable hosts
Component: Core & setup
Level: 1
Class: Bug fix
Version: 1.6.0i1
Hosts with a state differing from <tt>UP</tt> will no longer be considered in the
automatic check discovery routine.
ID: 6969
Title: Fixed recurring flexible downtimes.
Component: cmc
Level: 2
Class: Bug fix
Version: 1.6.0i1
Previously, the combination of "recurring" and "flexible" was broken for
downtimes. The effect was that such downtimes remained hot even outside
their intended time window, so that the next occurring problem just
triggered the start of such a downtime and no problem was notified.
An example for such a scenario: A flexible host downtime was configured to
happen every day between 02:00 and 03:00 for 2 hours. Everything was OK
between 02:00 and 03:00, but at 08:15 the host went DOWN. This started the
downtime, lasting until 10:15, which is of course totally unintended: The
problem did not happen between 02:00-03:00, so the downtime should not start
and normal problem processing should be done, including notifications etc.
This has been fixed, so the recurring flexible downtimes are working as
intended now. If you update your installation, there is nothing more you
have to do, all downtimes automatically work correctly after that. If you
do not want to update yet, you should delete your recurring flexible
downtimes for now and add new recurring fixed downtimes instead as a
workaround until you update.
ID: 7069
Title: agent_kubernetes: the kilo suffix for resources may be lowercase
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 1.6.0i1
Resolve an issue where parsing of memory failed if the suffix for
memory resources was 'k' instead of 'K'.
ID: 7068
Title: Truncate values of the notification context which are too long
Component: Notifications
Level: 1
Class: Bug fix
Version: 1.6.0i1
If the the value stored in the notification context was larger than
the OS limit MAX_ARG_STRLEN a failed notification was produced and
the error 'OSError: [Errno 7] Argument list too long' appeared in
the notify.log. This was usually the case for very long values in
the long outputs of a service like e.g. fileinfo.groups. Now the
values are truncated and the text 'Attention: Removed remaining
content because it was too long.' is shown at the end.
ID: 7221
Title: check_http: Use proxy settings for certificate check
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 1.6.0i1
Previously the proxy settings had been ignored when checking a certificate
unless the option "Check Certificate of different IP / DNS Name" had been used.