[Release] Checkmk stable release 1.6.0p12

Dear friends of Checkmk,

the new stable release 1.6.0p12 of Checkmk is ready for download.

With this release we publish the second feature pack of Checkmk. The idea
behind the feature packs is that we can make selected features of the next
stable release available to users of the current stable release much earlier.
This allows features to be used and improved earlier.

You can find more information about the feature pack here: https://checkmk.com/check_mk-werks.php?werk_id=10886

This maintenance release ships with 54 changes affecing all editions of Checkmk,
4 Enterprise Edition specific changes and 4 Managed Services Edition specific changes.

Changes in all Checkmk Editions:

Checks & agents:

  • 10886 Feature pack 2: Prometheus, RabbitMQ, Couchbase, Pulse Secure, Redis, Cisco Webex Teams, Cisco ASA
  • 10906 Extend UCD SNMP scan function
  • 10931 VMWare UAG (Photon OS) is using UCD SNMP now
  • 10152 chrony: Do not discover if chrony can’t connect but ntp section is present
  • 10951 fortigate_sslvpn: Addition of WATO configuration option for number of tunnels
  • 10885 SEC: check_sftp: Fix possible command injection from the WATO configuration
  • 11028 FIX: Fixed plugins README
  • 11136 FIX: Instance detection of mk_postgres
  • 10892 FIX: Make hr_mem work with corrupt memory units (e.g. ‘4096 Bytes’)
  • 10988 FIX: Periodic service discovery: Fix not working “rediscovery scheduled” on clusters
  • 11041 FIX: check_mailboxes: only use SSL encryption if configured
  • 11038 FIX: cifsmounts: fix erroneous display of ‘stale fs handle’
  • 11044 FIX: dell_idrac_fans: don’t crash on unknown device states
  • 10363 FIX: f5_bigip_mem: Check no longer crashes when memory information are missing
  • 10904 FIX: f5_bigip_mem: Fix crash on discovery
  • 11042 FIX: heartbeat_crm: fix vanished services for Pacemaker >= 2.0
  • 11104 FIX: heartbeat_crm: go to CRIT if data is too old
  • 10604 FIX: hyperv_vmstate: Consider Protocol_Mismatch ‘OK’
  • 11076 FIX: mk_docker.py: Do not create agent section when executed on non-docker host
  • 7885 FIX: nfsiostat: Fix broken configuration
  • 11040 FIX: nfsiostat: don’t crash if dots are present in mount point
  • 11103 FIX: systemd_units: Fix parsing error
  • 11123 FIX: ups_capacity can handle battery/time_left = 0
  • 11100 FIX: veeam_tapejobs: Fix crash upon deprecation warning
  • 11106 FIX: winperf_processor.util: Crash in case of services disovered pre 1.6
  • 11101 FIX: winperf_processor: Ineffective ruleset
  • 10778 FIX: Cisco CPU checks: relax SNMP scan function
    NOTE: Please refer to the migration notes!
  • 11018 FIX: mbg_lantime_ng_fan: Do not discover fans in state “not available”
    NOTE: Please refer to the migration notes!
  • 11005 FIX: veeam_client: Remove all non-printable ASCII characters from piggyback section name
    NOTE: Please refer to the migration notes!

Core & setup:

  • 10877 FIX: Fix “cmk” call with more than two -v arguments
  • 10883 FIX: Prevent empty service descriptions from being activated
  • 11016 FIX: Validate datasource of cluster and its nodes

Event console:

  • 11029 FIX: Prevent duplicated rule packs when updating mkp

HW/SW inventory:

  • 10987 FIX: HW/SW Inventory Service: Fix “State when inventory fails”


  • 11075 FIX: victorops: Fix invalid message_type (OK to RECOVERY)

Other components:

  • 11058 FIX: NagVis: Updated to 1.9.20

User interface:

  • 10887 FIX: Availability timeline: Escape service details equal to status views
  • 10878 FIX: Fix BI aggregation state export (Regression caused by #10879)
  • 11023 FIX: Fixed UnicodeDecodeError ‘ascii codec can’t decode byte’ when acknowledging log file if file name contains non-ASCII signs
  • 11014 FIX: Fixed possibility to delete roles that are still in use
  • 11015 FIX: Fixed renaming of roles
  • 11073 FIX: NagVis snapin: Fix empty content during refresh
  • 11020 FIX: Quicksearch: Validate if search string contains lookahead
  • 11022 FIX: Show only hosts in host dropdowns, the user is permitted for
  • 10889 FIX: Web API: site_login call did not work (regression since #6929)
  • 11021 FIX: Fixed validation of regex filter
    NOTE: Please refer to the migration notes!


  • 10535 Activate Changes: Now able to specifiy the number of parallel site activations
  • 10930 Auto monitor checkmks tmpfs during site creation
  • 10537 FIX: Activate changes: Ensure housekeeping, always cleanup obsolete files
  • 11026 FIX: Consider timespecific parameters for manual checks
  • 10779 FIX: Fix sorting by WATO folder
  • 11013 FIX: Fixed validation of datasource type for cluster and its nodes
  • 11027 FIX: LDAP: Validate usage of suffix
  • 10884 FIX: check_http: Add better validation for empty service names when using “^” prefix

Changes in the Checkmk Enterprise Edition:

Agent bakery:

  • 11012 FIX: Fixed unicode usage in fileinfo baklet

User interface:

  • 11079 FIX: Grafana: Annotation queries were not applying host filter


  • 10693 Agent Updater supports Windows UTF-8 encoding
  • 10782 FIX: Agent bakery generator for sap.cfg

Changes in the Checkmk Managed Services Edition:

User interface:

  • 10882 FIX: Fix “Customers” dashlet shown on customer sites
  • 11025 FIX: LDAP: Fixed default user attribute ‘customer’ on remote sites
  • 10413 FIX: Sites can not be assigned to global scope


  • 11074 FIX: Fix broken activate changes when EC config sync is enabled

You can download Checkmk from our download page:

Please mail bug reports and qualified feedback to feedback@checkmk.com.
We greatly thank you for using Checkmk and wish you a successful monitoring,

Your Checkmk Team


Great Release Notes!

Now if the Werk-IDs would be links to the actual werks, that would be the icing on the cake.


Yes, that would be convenient. As a workaround, we can use the werks overview page and filter for the branch+release in question.


1 Like

Hi All,

I noticed that the CMK Ubuntu 20.04 version was available and gave that a try:

  • upgraded the VM from Ubuntu 18.04 > 20.04
  • installed 1.6.0p12 CRE

Installing / upgrading worked ok, I used this page as reference during the upgrade to not lose check_mk config:

OS upgrade

Afterwards, check_mk starts polling, which al seems good, however I do get an error when viewing graphs. This seems related to this issue:


To me this looks like a bug, as al PHP versions are installed by the CMK package.

Are there other people on 20.04 with this same behaviour?
I raised this ticket : FEED-4973: Bug in Ubuntu 20.04 CRE | Function get_magic_quotes_runtime() is deprecated

Best regards,


The bug lies inside the Kohana framework and the used fpdf lib. Quick solution.
Edit the file
Starting at line 56 remove he two blocks

// magic_quotes_runtime is enabled
// magic_quotes_gpc is enabled

the second file is
starting from line 1042 remove the next two lines.

Graphs should work but PDF export from PNP4Nagios is broken because there are other errors inside the fpdf lib.

1 Like

Hi Andreas,

Thank you for the quick feedback, and getting our monitoring tool back alive.
Your suggestion resolves the graph problem.

Best regards,


I downloaded https://checkmk.com/support/1.6.0p12/check-mk-raw-1.6.0p12_0.focal_amd64.deb
from https://checkmk.com/download.php?edition=cre&version=stable&dist=ubuntu&os=focal
but always get a SHA256 hash of
instead of
which is shown on the download page.

Can someone please check this?



I can confirm your observation:

Apparently something is rotten in the state of Bavaria ¹ :roll_eyes:

Hi @Dirk, I have seen this problem also with the Debian 10.0 version. I’m not sure whats wrong: (a) the mentioned checksum on the download page or (b) the package.

Will wait until this is fixed before downloading and installing. Do you know if the responsibles are reading this or should I mail to the feedback address?

Thanks for your fast confirmation,

I’m not sure whether they will read this. Actually I doubt it because this is just a response to their announcement. But if we mention @Checkmk-Announcement then that account probably gets notified.

Another possibility might be to create a new post in the Feedback category or, as you suggest, send an email to the feedback address.


The problem is that inside the HASHES file there are more than one checksum per file.

This one is inside the file but also two other checksums for the same file :slight_smile:

d9a17a3ab68ffb29a7f6c544c8edda669c3dc0b3b5916828396271d249910b49  ./check-mk-raw-1.6.0p12_0.focal_amd64.deb
2dce436f35014600c4c48f779a0fd88db75f4107850a998f8c7290a0ac72aac7  ./check-mk-raw-1.6.0p12_0.focal_amd64.deb
be908f6e4fe0ac1c20d8082e8928de77a9960d8bc7dd6ccec0d7d39f63edbdb0  ./check-mk-raw-1.6.0p12_0.focal_amd64.deb

Hi guys, thanks for pointing us to this issue. It’s fixed now.

We had several rebuilds of this version that day and apparently the build system kept appending the hashes to the file Andreas mentioned above.

Sorry for the inconvenience.

1 Like

Works perfectly. Thanks for sharing the solution.

Hi everyone,



we have been using the cluster functionality quite often for services that have been put behind a Layer 4 load balancer (FortiADC with Direct Return) - and because we don’t monitor the load balancer this way, we configured the clusters with “No datasource” and the actual nodes with the Checmk agent.
Granted, we didn’t actually use any clustered services on those clusters, we just used it as a representation of our logical organisation.
This doesn’t work anymore now - what can we do? Just make the cluster objects regular hosts?


You should be fine by just giving the cluster also the datasource “Checkmk agent”?!

1 Like

We will try to implement that linking the werks directly.

Also, thanks everyone for sharing their feedback.

1 Like

Hi Andresa,

Any luck for pdf export issue.