ID: 6397
Title: Fixed too high CPU utilization on SNMP devices like HP procurve switches
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 1.6.0i1
After upgrading Check_MK from former than 1.5.0 some SNMP
devices like HP procurve core switches were reporting a
highly increased CPU utilization. This was caused by
SNMP check plugins like the network interface check plugin
being executed before the CPU utilization check.
ID: 6469
Title: New Theme: Fixed again several bugs
Component: Multisite
Level: 1
Class: Bug fix
Version: 1.6.0i1
This werk includes Fixes for:
- Hover with "loading..." message of Graph Preview
- Text shadow of PEND in services of hosts view
- Size of Perf-O-Meter in views where no headers are configured
- Missing line break in EC SNMP MIBs
- Size of column of acknowledge in analyze configuration if no icon is shown
- default template in reports is now in new colors, too
- Snapins in dashboards are now in new theme, too
- Date filter in Release notes
ID: 6450
Title: Fixed handling of "SNMP" and "Dual Check_MK Agent + SNMP" tags when migrating from <1.5
Component: WATO
Level: 2
Class: Bug fix
Version: 1.6.0i1
In addition to the problem that has been fixed with #6446, the werk #5535 introduced a problem when
one has not modified the tag group "Agent type".
In this situation the "Agent type" tag group is automatically changed to the
new format (as described in #5535).
During this process two tags were not migrated correctly:
<ul>
<li>Agent type: SNMP (Networking device, Appliance) (Tag-ID: snmp-only)</li>
<li>Agent type: Dual (Dual: Check_MK Agent + SNMP) (Tag-ID: snmp-tcp)</li>
</ul>
When editing the host, the new Agent type and SNMP attributes are selected correctly and also saved
correctly. No host lost his attributes in this case. But when saving the tag list of hosts to the
config files, the tags "snmp-only" and "snmp-tcp" were not set anymore.
This is a problem because existing rules may refer to these tags.
The second part of this problem are these probably existing rules. When you open them in the GUI and
saving them to the disk again, these rules will loose their conditions, because these tags did not
exist anymore.
We have now changed the migration procedure to work like this:
<ul>
<li>In the moment you update and the old "Agent type" tag group is migrated to the new one two new
auxiliary tags are added to your configuration (snmp-only: "Legacy: SNMP ..." and snmp-tcp: "Legacy: Dual:...")</li>
<li>Now editing of existing rules will work again</li>
<li>Hosts that have "Agent type" configured to "No agent" and "SNMP" to "SNMP v2 or v3" get
have the aux tag snmp-only set</li>
<li>Hosts that have "Agent type" configured to "Contact either Check_MK Agent or use datasource program"
and "SNMP" to "SNMP v2 or v3" will have the aux tag snmp-tcp set</li>
<li>Now the rules mentioned above will match again on these hosts</li>
<li>In case you want to clean up the legacy aux tags, you can now try to delete them. The GUI will warn
you about all rules referring to them. You'll now have to clean them up.</li>
<li>"Legacy: SNMP ..." needs to be converted to "SNMP -> SNMP v2 or v3" and "Agent: No Agent"</li>
<li>"Legacy: Dual: ..." needs to be converted to "SNMP -> SNMP v2 or v3" and "Agent: Contact either Check_MK Agent or use datasource program"</li>
<li>After cleaning up all refereces, you can delete the legacy tags.</li>
<ul>
ID: 6449
Title: Fixed stored XSS using custom host / user attributes
Component: Multisite
Level: 1
Class: Security fix
Version: 1.6.0i1
A user with admin privileges could inject arbitrary JS code into custom
attributes which could then be executed in the context of other users.
ID: 6448
Title: Fixed displaying tag aliases in rule lists for builtin tags
Component: Multisite
Level: 1
Class: Bug fix
Version: 1.6.0i1
Configured builtin tags and aux tags in conditions were displayed with their
tag ID instead of the alias.
ID: 6447
Title: Agent type: Fixed not working "use all enabled datasource programs" option
Component: Core & setup
Level: 1
Class: Bug fix
Version: 1.6.0i1
Werk #6556 introduced a new option "use all enabled datasource programs" that can
be used to disable the TCP Check_MK agent and use all special agents / data source
programs to fetch the monitoring data for a host.
This new option was not applied correctly. The hosts using this option were monitored
like they had the option "Contact either Check_MK agent or use data source program"
configured.
ID: 6016
Title: Fixed creation of rule-based downtimes
Component: cmc
Level: 2
Class: Bug fix
Version: 1.5.0p2
Due to a bug introduced in 1.5.0i3, it was impossible to create new working
rule-based downtimes: Although the corresponding rules seemed to work fine,
no actual downtimes were created from them, as could be seen from the
"Downtimes" view.
ID: 6446
Title: Fixed changing agent type from "No agent" to agent in some cases
Component: WATO
Level: 2
Class: Bug fix
Version: 1.6.0i1
With werk #5535 we changed the tag group "Agent type" to be a builtin tag
group. When a pre-1.5 site modified this tag group before updating to 1.5, this
tag group was left untouched.
In this situation hosts that were marked as "No agent" (ping only hosts) were
suddenly set to "Check_MK Agent (Server)" which resulted in error messages
in the monitoring after activating the configuration, because these hosts got
a Check_MK service and tried to contact the agent while there is none on the
affected hosts.
We have now changed the migration logic to extend the locally modified "Agent
type" tag group with the options Check_MK 1.5 needs. This change should make
your "No agent" hosts now stick with the intended configuration.
In case you are affected by this issue, you may notice that the choices of the
"Agent type" are now extended with some new options which may be confusing in
the first moment. You can now navigate to the "Host tags" page in WATO and edit
the tag group "agent" to sort the new options according to your needs.
The tag choices in the tag group "Agent type" that are available in your site
after the update (that are not part of the 1.5 standard) are displayed with the
"Legacy: " prefix. This can be useful in the moment you want to clean them up.
To migrate the legacy choices you need to clean up all objects referring these
choices before they can safely be removed. Examples:
<ul>
<li>The legacy "No Agent" choice was using the Tag ID "ping". The new one uses
the "no-agent" tag ID. You can now delete the legacy choice from the tag
group and hit save. The GUI will then inform you about all hosts, folders
or rulesets using this tag choice. Navigate to all objects and change the
old setting to the new one. After that you can safely remove the legacy
choice.</li>
<li>For the snmp-only, snmp-v1 and snmp-tcp options it's a bit more complicated
because it involves now two tag groups (SNMP + Agent type). To use the new
scheme you need to set the configuration like this:
<ul>
<li>snmp-only: Agent type = "No agent" and SNMP = SNMP v2 or v3</li>
<li>snmp-v1: Agent type = "No agent" and SNMP = SNMP v1</li>
<li>snmp-tcp: Agent type = "Contact either Check_MK Agent or use datasource program" and SNMP = SNMP v2 or v3</li>
</ul>
</li>
</ul>
ID: 5932
Title: New SNMP management board checks for HP, FSC and Dell devices
Component: Checks & agents
Level: 1
Class: New feature
Version: 1.5.0i4
If a TCP host has additionally a SNMP management board data source
in Check_MK this board discovers suitable SNMP checks.
But if there's a SNMP host which also has a SNMP management board
Check_MK must know which discovered check type belongs to which
data source. Therefore we have to create new checks for SNMP
management boards.
The following check types support SNMP management boards:
- mgmt_fsc_sc2_cpu_status
- mgmt_fsc_sc2_mem_status
- mgmt_fsc_sc2_fans
- mgmt_fsc_sc2_power_consumption
- mgmt_fsc_sc2_info
- mgmt_fsc_sc2_temp
- mgmt_fsc_sc2_voltage
- mgmt_dell_poweredge_cpu
- mgmt_dell_poweredge_mem
- mgmt_dell_poweredge_netdev
- mgmt_dell_poweredge_pci
- mgmt_dell_poweredge_status
- mgmt_dell_poweredge_psu
- mgmt_dell_poweredge_fans
- mgmt_dell_poweredge_amperage.power
- mgmt_dell_poweredge_temp
- mgmt_hp_proliant_da_cntlr
- mgmt_hp_proliant_cpu
- mgmt_hp_proliant_fans
- mgmt_hp_proliant_mem
- mgmt_hp_proliant_temp
- mgmt_snmp_info
ID: 6460
Title: jolokia_generic: Do not crash when non-numeric data is supplied for number or rate type values
Component: Checks & agents
Level: 1
Class: Bug fix
Version: 1.6.0i1
Previously, a single MBean sending a non-numeric value would crash all jolokia_generic checks on
the affected host. This has been fixed.