ID: 10469
Title: New plugin: Operations of Couchbase buckets
Component: Checks & agents
Level: 1
Class: New feature
Version: 1.7.0i1
A new plugin is introduced monitoring the Couchbase buckets' operations.
Please configure the Couchbase special agent to discover these services.
ID: 10380
Title: Fix view action menu in dashlets
Component: Multisite
Level: 1
Class: Bug fix
Version: 1.7.0i1
When clicking on the action dropdown menu of views embedded in dashlets, this
opened a new window. Clicking on "Reschedule ..." did also open a new window.
ID: 10381
Title: Fix editing dashlet views
Component: Multisite
Level: 1
Class: Bug fix
Version: 1.7.0i1
The general view options, painter, sorters and group painters could not be
edited once a view was added to a dashboard.
ID: 7284
Title: Introduced explicit host attributes, increasing rule evaluation performance
Component: WATO
Level: 1
Class: New feature
Version: 1.7.0i1
Some host attributes, like the host alias, are now stored in an explicit lookup table.
The previous version created a rule for each alias, hereby causing lots of uneccessary rule evaluations.
This is an internal code change, so you will not see any changes in the GUI.
ID: 10379
Title: Hostname search: Host bulk actions affected all hosts (1.6.0p4 regression)
Component: WATO
Level: 1
Class: Bug fix
Version: 1.7.0i1
When using the WATO host search to filter hosts by a host name pattern, the
resulting page showed the correctly matched hosts. When performing a bulk
operation on this page, this affected all hosts instead of the filtered hosts.
This regression was introduced by #10259 with 1.6.0p5.
ID: 10342
Title: HW/SW Inventory: Do not overwrite inventory tree if ALL data sources of a host fail
Component: HW/SW Inventory
Level: 1
Class: Bug fix
Version: 1.7.0i1
If a host has multiple data sources and ALL of these sources have an exception
like empty data, connection problems or timeouts then the inventory tree won't
be overwritten. This may reduce flapping inventory history entries.
If all sources fail the {{HW/SW Inventory}} service reports
{{Cannot update tree (WARN)}}.
Note:
This also concerns hosts with exactly one data source. If this source fails the
inventory tree won't be overwritten, too.
ID: 10402
Title: Deal with RRD sources which are not updated anymore in multiline graphs
Component: metrics
Level: 1
Class: Bug fix
Version: 1.7.0i1
The metrics "Guest operating systems" and "CPU time stolen by other
operating systems" that appear on the CPU utilization service Check for
unix systems are only relevant on virtualized environments. On other cases
they are always zero, and thus completely skipped in the graphs, but still
tracked by the monitoring core.
Checkmk 1.6 stopped delivering this metrics for systems where they are
irrelevant. This saves disk space, and disk I/O improving scalability.
If you update from Checkmk 1.5, the unused RRDs will remain in disk but
won't be updated anymore. If you use the RRD single format you wont
experience any problem, but if you are on RRD format for one RRD per
metric, the RRDs will go out of sync. At the time of drawing the graphs,
this lack of syncronicity will disrupt the resolution of the data, and
crash the graphs, if two metrics have a different resolution.
With this werk, data is resampled after being retrieved from the core if
the metrics on the same graph have different resolutions.
ID: 10422
Title: cisco_cpu_memory: Check for monitoring CPU Memory of Cisco devices
Component: Checks & agents
Level: 1
Class: New feature
Version: 1.7.0i1
This check monitors the CPU Memory level of Cisco devices. A Wato rule can be configured
depending on free or occupied memory. The details displayed on the service depend on this
WATO rule with the default being free memory. The performance graph will always be based
on occupied memory.
ID: 10360
Title: agent_aws: Fetch live data from AWS if special agent configuration has changed
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 1.7.0i1
The special agent {{agent_aws}} fetches data from an AWS account relying on the
WATO configuration ruleset {{Amazon Web Services (AWS)}} for a given host.
Cloudwatch metrics and other metrics have a resolution of 5 minutes by default.
(S3 has 1 day). In order to save API calls and costs the agent caches the data.
If the special agent configuration changes the agent does not recognize that
and still uses old cached data until the cache expires.
Now the special agent computes a hash over the current configuration and if the
configuration changes the agent fetches live data immediately.
The configuration hash will be stored in
{{tmp/check_mk/agents/agent_aws/HOSTNAME.config_hash}}.
>From {{cmk -D HOSTNAME}} you get the commandline of the special agent for a
given host. With the additional option {{--no-cache}} you can update the cached
data, too.
Workaround for previous special agent versions:
Remove the files {{tmp/check_mk/agents/agent_aws/REGION/HOSTNAME/*}}.