ID: 14890
Title: Fix possible error on filtering service level of hosts
Component: Multisite
Level: 1
Class: Bug fix
Version: 2.2.0i1
Using the filter "Host service level" for views with hosts with configured
custom attributes could lead to an error message like "Internal error: invalid
literal for int() with base 10".
ID: 14563
Title: <tt>heartbeat_rscstatus<\tt>: Fix error during update
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 2.2.0i1
During discovery the plugin <tt>heartbeat_rscstatus<\tt> stores its discovery state. Updating
Checkmk from 2.0.0 to 2.1.0 or updates within those versions would cause these states to be lost.
This is indicated by following error message:
C+:
-| 8/27 Rewriting autochecks...
-| Transform failed: host='host', plugin='heartbeat_rscstatus', ruleset='heartbeat_rscstatus', params={'discovered_state': 'none'}, error=AssertionError('non-empty params vanished')
C-:
With this werk, the states are correctly persisted.
ID: 14300
Title: Make appliance snapin compatible with Checkmk Appliance firmware 1.6+
Component: Multisite
Level: 1
Class: New feature
Version: 2.2.0i1
When Checkmk is used on on the Checkmk Appliance 1.6 or newer, it is necessary
to update the appliance sidebar snapin.
Previous Checkmk versions displayed the error message 'Failed to render navigation:
Traceback (most recent call last): File: "[stdin]", line 2, in <module>
IOError: [Errno 2] No such file or directory' in the sidebar snapin.
Background: The Checkmk Appliance firmware 1.6 migrates the internal software
stack to Python 3.7 which makes it incompatible with the previous snapin. This
change makes it possible to use Checkmk on older and newer appliance firmware.
ID: 14901
Title: user_config: REST API fix to ensure the request and response schemas align
Component: REST API
Level: 1
Class: Bug fix
Version: 2.2.0i1
This werk introduces a fix that aligns both the response and request schemas on the user_config
endpoints. The response schema now nests the enforce_password_change attribute just like the
request schemas. So the response from a GET request has the same format as what is required in
a POST/PUT request. Client scripts should be adjusted accordingly.
ID: 14896
Title: Fix "Move to other folder" of hosts filtered by alias
Component: Setup
Level: 1
Class: Bug fix
Version: 2.2.0i1
If you filtered for hosts on the "Setup" - "Hosts" page and used the alias as
filter value, the option "Move to other folder" showed the error "Please select
some hosts before doing bulk operations on hosts.", even if you selected the
matched hosts.
ID: 14655
Title: Extension manager: disable upgraded packages
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 2.2.0i1
When upgrading installed MKPs to a new version, the former version is disabled now.
ID: 13262
Title: Missing failure indicator in IPMI status messages
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 2.2.0i1
The IPMI sensor check would result in an OK state when confronted with a status
message like this:
"ok (Presence detected, Predictive failure, Power Supply AC lost)"
This change adds more substrings indicating an error and renders a service with
the above string CRIT.
ID: 14958
Title: cmk-update-agent: Fix lockfile handling
Component: agents
Level: 1
Class: Bug fix
Version: 2.2.0i1
This Werk mainly fixes an inconvenience that occured when invoking multiple
instances of the Agent Updater simultaneously.
Since Werk #14732, it's possible for the Agent Updater to wait for running exclusively.
However, before this Werk, a second instance failed to actually open and hold the
respective lockfile (because the first instance deleted it). Hence, it also failed to
delete it in the end.
This resulted in an error message like <tt>No such file or directory:
'/tmp/cmk-update-agent.pid'"</tt>. This didn't result in a crash, but the error was
printed to the logfile and could lead to confusion.
ID: 14957
Title: fileinfo: Crash on negative timespans
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 2.2.0i1
When monitoring files/filegroups with service monitoring rules <i>Size and age of single files</i>
or <i>Size, age and count of file groups</i> (i.e., all services basing on the <i>fileinfo</i>
agent section, not on the mk_filestats agent plugin), the service evaluation could fail with a crash
report containing the error "Cannot render negative timespan".
This happened very rarely, and if it did, it was temorary (for one check period).
Since the fix consists of a change in the agent section evaluation, you'll have to update agents
to apply it on affected systems.