MPF.md

The Masterfiles Policy Framework is the default policy that ships with both the CFEngine Enterprise and Community editions. The MPF includes policy to manage cfengine itself, the stdlib, and policy to inventory various aspects of the system.

Overview

• update.cf - The update policy entry.
• promises.cf - The main policy entry.
• lib/ - The standard library.
• inventory/ - Inventory policy.
• services - User defined custom policy.
• services/main.cf - Contains an empty bundle agent main where custom policy can be integrated.
• services/autorun/ - Automatically included policy files.
• .no-distrib/ - A directory that is excluded from policy updates from remote agents.

The MPF is continually updated. You can track its development on github.

Configuration

The most common controls reference variables in the def bundle in order to keep the modifications to the distributed policy contained within a single place. The recommended way to set classes and define variables in the def bundle is using an [augments file][Augments]. Keeping the modifications to the distributed policy set makes policy framework upgrades significantly easier.

Note: If you need to make modification to a shipped file consider opening a pull request to expose the tunable into the def bundle.

Note: controls/def.cf contains the defaults and settings for promises.cf and controls/update_def.cf contains the defaults and settings for update.cf.

Since 3.7.8, 3.10.4, and 3.12.0 The class cf_runagent_initated is defined by default in the MPF for agent executions initiated by cf-runagent through cf-serverd. Previously the class cfruncommand was defined. See body server control cfruncommand in controls/cf_serverd.cf.

Update Policy (update.cf)

Synchronizing clients with the policy server happens here, in update.cf. Its main job is to copy all the files on the policy server (usually the hub) under $(sys.masterdir) (usually /var/cfengine/masterfiles) to the local host into $(sys.inputdir) (usually /var/cfengine/inputs).

This file should rarely if ever change. Should you ever change it (or when you upgrade CFEngine), take special care to ensure the old and the new CFEngine can parse and execute this file successfully. If not, you risk losing control of your system (that is, if CFEngine cannot successfully execute update.cf, it has no mechanism for distributing new policy files).

By default, the policy defined in update.cf is executed at the beginning of a cf-execd scheduled agent run (see schedule and exec_command as defined in body executor control in controls/cf_execd.cf). When the update policy completes (regardless of success or failure) the policy defined in promises.cf is activated.

This is a standalone policy file. You can actually run it with cf-agent -KI -f ./update.cf but if you don't understand what that command does, please hold off until you've gone through the CFEngine documentation. The contents of update.cf duplicate other things under lib sometimes, in order to be completely standalone.

To repeat, when update.cf is broken, things go bonkers. CFEngine will try to run a backup failsafe.cf you can find in the C core under libpromises/failsafe.cf (that .cf file is written into the C code and can't be modified). If things get to that point, you probably have to look at why corrupted policies made it into production.

As is typical for CFEngine, the policy and the configuration are mixed. In controls/update_def.cf you'll find some very useful settings. Keep referring to controls/update_def.cf as you read this. We are skipping the nonessential ones.

Configure upstream masterfiles location for policy update

Want to get your policy from a place other than /var/cfengine/masterfiles on sys.policy_hub?

With an augments like this:

{
  "vars": {
            "house": "Gryffindor",
            "mpf_update_policy_master_location": "/srv/cfengine/$(sys.flavor)/$(def.house)"
          }
}

A CentOS 7 host would copy policy from /srv/cfengine/centos_6/Gryffindor to $(sys.inputdir) (commonly /var/cfengine/inputs).

History:

• Introduced in 3.12.0.

Append to inputs used by update policy

You can append to the inputs used by the update policy via augments by defining vars.update_inputs. The following example will add the policy file my_updatebundle1.cf to the list of policy file inputs during the update policy.

{
  "vars":{
    "update_inputs": [ "my_updatebundle1.cf" ]
  }
}

Append to the update bundlesequence

You can specify bundles which should be run at the end of the default update policy bundlesequence by defining control_common_update_bundlesequence_end in the vars of an [augments file][Augments].

For example:

{
  "vars":{
    "control_common_update_bundlesequence_end": [ "my_updatebundle1", "mybundle2" ]
  }
}

Notes:

• The order in which bundles are actuates is not guaranteed.
• The agent will error if a named bundle is not part of inputs.

Specify the agent bundle used for policy update

The MPF uses cfe_internal_update_policy_cpv to update inputs and modules on remote agents. When new policy is verified by the agent /var/cfengine/masterfiles/cf_promises_validated is updated with the current timestamp. This file is used by remote agents to avoid unnecessary inspection of all files each time the update policy is triggered.

Override this bundle by setting def.mpf_update_policy_bundle via augments:

{
    "vars": {
        "mpf_update_policy_bundle": "MyCustomPolicyUpdateBundle"
    }
}

History:

• Introduced in 3.12.0

ignore_missing_bundles

This option allows you to ignore errors when a bundle specified in body common control bundlesequence is not found.

This example illustrates enabling the option via augments.

{
  "vars": {
    "control_common_ignore_missing_bundles": "true"
  }
}

NOTE: The same augments key is used for both update.cf and promsies.cf entries.

History:

• Introduced in 3.12.0

ignore_missing_inputs

This option allows you to ignore errors when a file specified in body common control inputs is not found.

This example illustrates enabling the option via augments.

{
  "vars": {
    "control_common_ignore_missing_inputs": "true"
  }
}

NOTE: The same augments key is used for both update.cf and promsies.cf entries.

History:

• Introduced in 3.12.0

Verify update transfers

Enable additional verrification after file transfers during policy update by defining the class cfengine_internal_verify_update_transfers. When this class is defined, the update policy will hash the transfered file and compare it against the hash given by the server

This [augments file][Augments] will enable this behavior for all clients.

{
  "classes": {
    "cfengine_internal_verify_update_transfers": [ "any" ]
  }
}

Encrypted transfers

Note: When using protocol version 2 or greater all communications are encapsulated within a TLS session. This configuration option is only relevant for clients using protocol version 1 (default for versions 3.6 and prior).

To enable encryption during policy updates define the class cfengine_internal_encrypt_transfers.

Preserve permissions

By default the MPF does not enforce permissions of inputs. When masterfiles are copied to inputs, new files are created with default restrictive permissions. If the class cfengine_internal_preserve_permissions is defined the permissions of the policy server's masterfiles will be preserved when they are copied.

Enable CFEngine Enterprise HA

When the enable_cfengine_enterprise_hub_ha class is defined the policy to manage High Availability of Enterprise Hubs is enabled.

Note: This class is not defined by default.

Disable cf_promises_validated check

For non policy hubs the default update policy only performs a full scan of masterfiles if cf_promises_validated is repaired. This repair indicates that the hub has validated new policy that the client needs to refresh.

To disable this check define the cfengine_internal_disable_cf_promises_validated class.

It not recommended to disable this check as it both removes a safety mechanism that checks for policy to be valid before allowing clients to download updates, and the increased load on the hub will affect scalability.

If you want to periodically perform a full scan consider adding custom policy to simply remove $(sys.inputdir)/cf_promises_validated. This will cause the file to be repaired during the next update run triggering a full scan.

Disable automatically removing files not present upstream (SYNC masterfiles)

By default, the MPF will keep inputdir in sync with masterfiles on the hub. If the class cfengine_internal_purge_policies_disabled is defined the update behavior will only keep files that exist on the remote up to date locally, files that exist locally that do not exist upstream will be left behind. Note, if this is disabled and a policy file that is dynamically loaded based on it's presence is renamed, duplicate definition errors may occur, preventing policy execution.

This [augments file][Augments] will enable this behavior for all clients.

{
  "classes": {
    "cfengine_internal_purge_policies_disabled": [ "any" ]
  }
}

History:

• Introduced in 3.18.0, previously, the default behavior was opposite and the class cfengine_internal_purge_policies had to be enabled to keep inputs in sync with masterfiles.

Disable limiting robot agents

By default the MPF (Masterfiles Policy Framework) contains active policy that is intended to remediate a pathological condition where multiple agent component daemons (like cf-execd) are running concurrently.

Define the class mpf_disable_cfe_internal_limit_robot_agents to disable this automatic remediation.

{
  "classes": {
    "mpf_disable_cfe_internal_limit_robot_agents": [ "any" ]
  }
}

History:

• Introduced in 3.15.0, 3.12.3, 3.10.7

Automatically deploy masterfiles from Version Control

On a CFEngine Enterprise Hub during the update policy if the class cfengine_internal_masterfiles_update is defined masterfiles will be automatically deployed from an upstream version control repository using the [settings defined via Mission Portal][Best Practices#Version Control and Configuration Policy] or directly in /opt/cfengine/dc-scripts.

Note: Any policy in the distribution location (/var/cfengine/masterfiles) will be deleted the first time this tooling runs. Be wary of local modifications before enabling.

Policy Analyzer Exclude Files

When the policy analyzer is enabled, a copy of the policy is made available for viewing from Mission Portal. To exclude files from this view you can define def.cfengine_enterprise_policy_analyzer_exclude_files as a list of regular expressions matching files that you do not want to be viewable from Policy Analyzer.

This [augments file][Augments] will prevent any files named please-no-copy and any file names that contain no-copy-me from being copied and visible from Policy Analyzer.

{
  "vars": {
    "cfengine_enterprise_policy_analyzer_exclude_files": [ "please-no-copy", ".*no-copy-me.*" ]
  },
}

History:

• Added in 3.19.0, 3.18.1

Policy Permissions

By default the policy enforces permissions of 0600 meaning that inputs are only readable by their owner. If you are distributing scripts with your masterfiles, be sure there is a policy to ensure they are executable when you expect them to be.

Agent binary upgrades

Remote agents can upgrade their own binaries using the built in binary upgrade policy. Packages must be placed in /var/cfengine/master_software_updates in the appropriate platform directory. Clients will automatically download and install packages when the trigger_upgrade class is defined during a run of update.cf.

This [augments file][Augments] will defines trigger_upgrade on hosts with IPv4 addresses in 192.0.2.0/24 or 203.0.113.0/24 or hosts that have a cfengine 3.10 class not for cfengine 3.10.2.

{
   "classes": {
    "trigger_upgrade": [
      "ipv4_10_10_1",
      "ipv4_10_10_2",
      "cfengine_3_10_(?!2$)\\d+"
    ]
   }
}

Notes:

• This policy is specific to CFEngine Enterprise.
• The negative look ahead regular expression is useful because it automatically turns off on hosts after they reach the target version.

Configure path that software is served from for autonomous agent upgrades

{% comment %}ENT-4953{% endcomment %} def.master_software_updates defines the path that cfengine policy servers share software updates from. Remote agents access this path via the master_software_updates shortcut. By default this path is $(sys.workdir)/master_software_updates. This path can be overridden via vars.dir_master_software_updates in augments.

For example:

{
   "vars": {
     "dir_master_software_updates": "/srv/cfengine-software-updates/"
   }
}

History:

• Introduced 3.15.0, 3.12.3, 3.10.8

Disable seeding binaries on hub

By default when trigger_upgrade is defined on a hub, the hub will download packages for agents to use during self upgrade. This automatic download behavior is disabled when the class mpf_disable_hub_masterfiles_software_update_seed is defined.

For example:

{
   "classes": {
     "mpf_disable_hub_masterfiles_software_update_seed": [ "policy_server::" ]
   }
}

History:

• Introduced 3.19.0, 3.18.1

Files considered for copy during policy updates

The default update policy only copies files that match regular expressions listed in def.input_name_patterns.

This [augments file][Augments] ensures that only files ending in .cf, .dat, .mustache, .json, .yaml and the file cf_promises_release_id will be considered by the default update policy.

{
    "vars": {
        "input_name_patterns": [ ".*\\.cf", ".*\\.dat", ".*\\.txt", ".*\\.conf",
            ".*\\.mustache", ".*\\.sh", ".*\\.pl", ".*\\.py", ".*\\.rb",
            ".*\\.sed", ".*\\.awk", "cf_promises_release_id", ".*\\.json",
            ".*\\.yaml", ".*\\.csv" ]
    }
}

Note: This filter does not apply to bootstrap operations. During bootstrap the embedded failsafe policy is used and it decides which files should be copied.

Configuring component management

The Masterfiles Policy Framework ships with policy to manage the components of CFEngine.

By default, for hosts without systemd, this policy defaults to ensuring that components are running.

On systemd hosts, the policy to manage component units is disabled by default.

Enable management of components on systemd hosts

To allow the Masterfiles Policy Framework to actively manage cfengine systemd units and state define the mpf_enable_cfengine_systemd_component_management.

This example illustrates enabling management of components on systemd hosts having a class matching redhat_8 via augments.

{
  "classes": {
    "mpf_enable_cfengine_systemd_component_management": [ "redhat_8" ]
  }
}

When enabled, the policy will render systemd unit files in /etc/systemd/system for managed services. Mustache templates for service units are in the templates directory in the root of the Masterfiles Policy Framework.

When enabled, the policy will make sure that all units are enabled, unless they have been disabled by a persistent class or are explicitly listed as an agent to be disabled.

Enable or disable CFEngine components

Using persistent classes

persistent_disable_DAEMON

Description: Disable a CFEngine Enterprise daemon component persistently.

DAEMON can be one of cf_execd, cf_monitord or cf_serverd.

This will stop the AGENT from starting automatically.

This [augments file][Augments] will ensure that cf-monitord is disabled on hosts that have server1 or the redhat class defined.

{
  "classes": {
    "persistent_disable_cf_monitord": [ "server1", "redhat" ]
  }
}

clear_persistent_disable_DAEMON

Description: Re-enable a previously disabled CFEngine Enterprise daemon component.

DAEMON can be one of cf_execd, cf_monitord or cf_serverd.

This [augments file][Augments] will ensure that cf-monitord is not disabled on redhat hosts.

{
  "classes": {
    "clear_persistent_disable_cf_monitord": [ "redhat" ]
  }
}

Using augments

agents_to_be_disabled

Description: list of agents to disable.

This [augments file][Augments] is a way to specify that cf-monitord should be disabled on all hosts.

{
  "vars": {
    "agents_to_be_disabled": [ "cf-monitord" ]
  }
}

Main Policy (promises.cf)

The following settings are defined in controls/def.cf can be set from an [augments file][Augments].

dmidecode inventory

When dmidecode is present, some key system attributes are inventoried. The inventoried attributes can be overridden by defining def.cfe_autorun_inventory_demidecode[dmidefs] via augments. dmidecode queries each key in dmidefs and tags the result with the value prefixed with attribute_name= Note, as the dmidefs are overridden, you must supply all desired inventory attributes.

For example:

{
  "vars": {
    "cfe_autorun_inventory_dmidecode": {
      "dmidefs": {
        "bios-vendor": "BIOS vendor",
        "bios-version": "BIOS version",
        "system-serial-number": "System serial number",
        "system-manufacturer": "System manufacturer",
        "system-version": "System version",
        "system-product-name": "System product name",
        "bios-release-date": "BIOS release date",
        "chassis-serial-number": "Chassis serial number",
        "chassis-asset-tag": "Chassis asset tag",
        "baseboard-asset-tag": "Baseboard asset tag"
      }
    }
  }
}

History:

• Introduced 3.13.0, 3.12.1, 3.10.5

Configure proc inventory

By default the MPF inventories consoles, cpuinfo, modules, partitions, and version from /proc. This can be adjusted by defining default:cfe_autorun_inventory_proc.basefiles.

For example:

{
  "variables": {
    "default:cfe_autorun_inventory_proc.basefiles": {
      "value": [
        "consoles",
        "cpuinfo",
        "version"
      ],
      "comment": "We do not need the extra variables this produces since we get the info differently",
      "tags": [
        "Custom override MPF default"
      ]
    }
  }
}

History:

• Added 3.21.0

Configure cf-agent syslog facility

To configure the syslog facility used by cf-agent configure agentfacility by setting default:def.control_agent_agentfacility via augments to one of the allowed values (LOG_USER, LOG_DAEMON, LOG_LOCAL0, LOG_LOCAL1, LOG_LOCAL2, LOG_LOCAL3, LOG_LOCAL4, LOG_LOCAL5, LOG_LOCAL6, LOG_LOCAL7)

{
  "variables": {
    "default:def.control_agent_agentfacility": {
      "value": "LOG_USER"
    }
  }
}

History:

• Added in 3.22.0

mailto

The address that cf-execd should email agent output to.

mailfrom

The address that output mailed from cf-execd should come from.

smtpserver

The SMTP server that cf-execd should use to send emails.

mailmaxlines

The maximumm number of lines of output that cf-execd will email.

History:

• Added in CFEngine 3.22.0, 3.21.1, 3.18.4

See also: [maxmaillines][cf-execd#maxmaillines]

Configure subject for emails sent by cf-execd

When enabled cf-execd emails output that differs from previous executions. The subject of the email can be configured by setting mailsubject in body executor control. This will use the value of default:def.control_executor_mailsubject if it is a non-empty string.

{
  "variables": {
    "default:def.control_executor_mailsubject": {
        "value": "CFEngine output from $(sys.fqhost)"
    }
  }
}

History:

• Added in 3.22.0

Configure lines that should be excluded from emails sent by cf-execd

When enabled cf-execd emails output that differs from previous executions. Lines matching regular expressions in mailfilter_exclude in body executor control are stripped before sending. The MPF will use the value of default:def.control_executor_mailfilter_exclude if it is a non-empty list.

{
  "variables": {
    "default:def.control_executor_mailfilter_exclude": {
        "value": [ ".*ps output line.*", ".*regline.*" ]
    }
  }
}

History:

• Added in 3.22.0

Configure lines that should be included from emails sent by cf-execd

When enabled cf-execd emails output that differs from previous executions. Lines matching regular expressions in mailfilter_include in body executor control are stripped before sending. The MPF will use the value of default:def.control_executor_mailfilter_include if it is a non-empty list.

{
  "variables": {
    "default:def.control_executor_mailfilter_include": {
        "value": [ ".*EMAIL.*" ]
    }
  }
}

History:

• Added in 3.22.0

acl

def.acl is a list of of network ranges that should be allowed to connect to cf-serverd. It is also used in the default access promises to allow hosts access to policy and modules that should be distributed.

Here is an example setting the acl from augments:

{
  "vars": {
    "acl": [ "24.124.0.0/16", "192.168.33.0/24" ]
  }
}

See Also: Configure networks allowed to make collect calls (client initiated reporting)

ignore_missing_bundles

This option allows you to ignore errors when a bundle specified in body common control bundlesequence is not found.

This example illustrates enabling the option via augments.

{
  "vars": {
    "control_common_ignore_missing_bundles": "true"
  }
}

NOTE: The same augments key is used for both update.cf and promises.cf entries.

History:

• Introduced in 3.12.0

ignore_missing_inputs

This option allows you to ignore errors when a file specified in body common control inputs is not found.

This example illustrates enabling the option via augments.

{
  "vars": {
    "control_common_ignore_missing_inputs": "true"
  }
}

NOTE: The same augments key is used for both update.cf and promises.cf entries.

History:

• Introduced in 3.12.0

trustkeysfrom

The list of network ranges that cf-serverd should trust keys from. This is should only be open on policy servers while new hosts are expected to be bootstrapped. It should be empty after your hosts have been bootstrapped to avoid unwanted hosts from being able to bootstrap.

By default the MPF configures cf-serverd to trust keys from any host. This is convenient for simplified bootstrapping. After initial deployment it is recommended that this setting be reviewed and adjusted appropriately according to the needs of your infrastructure.

The [augments][Augments] file (def.json) can be used to override the default setting. For example it can be restricted to 127.0.0.1 to prevent keys from any foreign host from being automatically accepted.

{
  "vars": {
    "trustkeysfrom": [ "127.0.0.1" ]
    }
}

Prevent automatic trust for any host by specifying an empty value:

{
  "vars": {
    "trustkeysfrom": [ "" ]
    }
}

Append to inputs used by main policy

The inputs key in augments can be used to add additional custom policy files.

See Also: [Append to inputs used by update policy][Append to inputs used by update policy]

History:

• Introduced in CFEngine 3.7.3, 3.12.0

services_autorun

See the documentation in [services/autorun][mpf-services-autorun].

postgresql_full_maintenance

On CFEngine Enterprise policy hubs this class is defined by default on Sundays at 2am. To adjust when postgres maintenance operations run edit controls/def.cf directly.

postgresql_vacuum

On CFEngine Enterprise policy hubs this class is defined by default at 2am when postgresql_maintenance_supported is defined except for Sundays.

To adjust when postgres maintenance operations run edit controls/def.cf directly.

enable_cfengine_enterprise_hub_ha

Set this class when you want to enable the CFEngine Enterprise HA policies.

This class can be defined by an [augments file][Augments]. For example:

{
  "classes" {
    "enable_cfengine_enterprise_hub_ha": [ "hub001" ]
  }
}

enable_cfe_internal_cleanup_agent_reports

This class enables policy that cleans up report diffs when they exceed def.max_client_history_size. By default is is off unless a CFEngine Enterprise agent is detected.

Configure splaytime

splaytime is the maximum number of minutes exec_commad should wait before executing.

Note: splaytime should be less than the scheduled interval plus agent run time. So for example if your agent run time is over 1 minute and you are running the default execution schedule of 5 mintues your splay time should be set to 3.

Configure it via augments by defining control_executor_splaytime:

{
  "vars": {
    "control_executor_splaytime": "3"
  }
}

Configure agent expiration

cf-agents spawned by cf-execd are killed after this number of minutes of not returning data.

Example configuration via augments:

{
  "vars": {
    "control_executor_agent_expireafter": "15"
  }
}

Configure agent execution schedule

The execution scheduled is expressesd as a list of classes. If any of the classes are defined when cf-execd wakes up then exec_command is triggered. By default this is set to a list of time based classes for every 5th minute. This results in a 5 minute execution schedule.

Example configuration via augments:

{
  "vars": {
    "control_executor_schedule": [ "Min00", "Min30" ]
  }
}

The above configuration would result in exec_command being triggered at the top and half hour and sleeping for up to splaytime before agent execution.

Configure cf-execd runagent socket users

On Enterprise hubs, access to cf-execd sockets can be configured as a list of users who should be allowed by defining vars.control_executor_runagent_socket_allow_users. By default on Enterprise Hubs, cfapache is allowed to access runagent sockets.

{
  "vars": {
    "control_executor_runagent_socket_allow_users": [ "cfapache", "vpodzime" ]
  }
}

History:

• Added in CFEngine 3.18.0

Allow connections from the classic/legacy protocol

By default since 3.9.0 cf-serverd disallows connections from the classic protocol by default. To allow clients using the legacy protocol (versions prior to 3.7.0 by default) define control_server_allowlegacyconnects as a list of networks.

Example definition in augments file:

{
  "vars": {
    "control_server_allowlegacyconnects": [ "0.0.0.0/0" ]
  }
}

Configure the ciphers which are allowed to be used by cf-serverd

When default:def.control_server_allowciphers is defined cf-serverd will use the ciphers specified instead of the binary defaults.

Example definition in augments file:

{
  "variables": {
    "default:def.control_server_allowciphers": {
     "value": "AES256-GCM-SHA384:AES256-SHA",
     "comment": "Restrict the ciphers that cf-serverd is allowed to use for better security"
    }
  }
}

Notes:

• Be careful changing this setting. A setting that is not well aligned between all clients and the server could result in clients not being able to communicate with the hub preventing further policy updates.

History:

• Added in 3.22.0

Configure the ciphers which are allowed to be used by cf-agent

When default:def.control_common_tls_ciphers is defined cf-agent will use the ciphers specified instead of the binary defaults for outgoing connections.

Example definition in augments file:

{
  "variables": {
    "default:def.control_common_tls_ciphers": {
     "value": "AES256-GCM-SHA384:AES256-SHA",
     "comment": "Restrict the ciphers that are used for outgoing connections."
    }
  }
}

Notes:

• Be careful changing this setting. A setting that is not well aligned between all clients and the server could result in clients not being able to communicate with the hub preventing further policy updates.
• This setting is instrumented in all of the default entry points (promises.cf, update.cf, standalone_self_upgrade.cf).

History:

• Added in 3.22.0

Configure the minimum TLS version which is allowed to be used by cf-serverd

When default:def.control_server_allowtlsversion is defined cf-serverd will use the minimum TLS version specified instead of the binary defaults.

Example definition in augments file:

{
  "variables": {
    "default:def.control_server_allowtlsversion": {
     "value": "1.0",
     "comment": "We want to allow old (<3.7.0) clients to connect."
    }
  }
}

Notes:

• Be careful changing this setting. A setting that is not well aligned between all clients and the server could result in clients not being able to communicate with the hub preventing further policy updates.

History:

• Added in 3.22.0

Configure the minimum TLS version which is allowed to be used by cf-agent

When default:def.control_common_tls_min_version is defined cf-agent will use the minimum TLS version specified instead of the binary defaults for outgoing connections.

Example definition in augments file:

{
  "variables": {
    "default:def.control_common_tls_min_version": {
     "value": "1.0",
     "comment": "We want to connect to old (<3.7.0) servers."
    }
  }
}

Notes:

• Be careful changing this setting. A setting that is not well aligned between all clients and the server could result in clients not being able to communicate with the hub preventing further policy updates.
• This setting is instrumented in all of the default entry points (promises.cf, update.cf, standalone_self_upgrade.cf).

History:

• Added in 3.22.0

Configure users allowed to initate execution via cf-runagent

cf-serverd only allows specified users to request unscheduled execution remotely via cf-runagent.

By default the MPF allows root to request unscheduled execution of non policy servers and does not allow any users to request unscheduled execution from policy servers.

To configure the list of users allowed to request unscheduled execution define vars.control_server_allowusers.

{
  "vars": {
    "control_server_allowusers": [ "root", "nickanderson", "cfapache" ],
  }
}

It's possible to configure different users that are allowed for policy servers versus non policy servers via vars.control_server_allowusers_non_policy_server and vars.control_server_allowusers_policy_server. However, if vars.control_server_allowusers is defined, it has precedence.

This example allows the users hubmanager and cfoperator to request unscheduled execution from policy servers and no users are allowed to request unscheduled runs from non policy servers.

{
  "vars": {
    "control_server_allowusers_non_policy_server": [ ],
    "control_server_allowusers_policy_server": [ "hubmanager", "cfoperator" ]
  }
}

History:

• Added in 3.13.0, 3.12.1
• Added vars.control_server_allowusers in 3.18.0

See Also: [Configure hosts allowed to initate execution via cf-runagent][Configure hosts allowed to initate execution via cf-runagent]

Configure hosts allowed to initate execution via cf-runagent

cf-serverd only allows specified hosts to request unscheduled execution remotely via cf-runagent.

By default the MPF allows policy servers (as defined by def.policy-servers) to initiate agent runs via cf-runagent.

To configure the list of hosts allowed to request unscheduled execution define vars.mpf_admit_cf_runagnet_shell. This example allows the IPv4 address 192.168.42.10, the host bastion.example.com, and the host with identity SHA=43c979e264924d0b4a2d3b568d71ab8c768ef63487670f2c51cd85e8cec63834 and policy servers the ability to initiate agent runs via cf-runagent.

{
    "vars": {
        "mpf_admit_cf_runagent_shell": [ "192.168.42.10",
                                         "bastion.example.com",
                                         "SHA=43c979e264924d0b4a2d3b568d71ab8c768ef63487670f2c51cd85e8cec63834",
                                         "@(def.policy_servers)"
                                       ]
    }
}

See Also: [Configure users allowed to initate execution via cf-runagent][Configure users allowed to initate execution via cf-runagent]

History:

• Added in CFEngine 3.18.0

Configure retention for files in log directories

By default the MPF rotates managed log files when they reach 1M in size. To configure this limit via augments define vars.mpf_log_file_max_size.

For example:

{
  "vars": {
    "mpf_log_file_max_size": "10M"
  }
}

By default the MPF keeps up to 10 rotated log files. To configure this limit via augments define vars.mpf_log_file_retention.

For example:

{
  "vars": {
    "mpf_log_file_retention": "5"
  }
}

By default the MPF retains log files in log directories (outputs, reports and application logs in Enterprise) for 30 days. This can be configured by setting vars.mpf_log_dir_retention via augments.

For example:

{
  "vars": {
    "mpf_log_dir_retention": "7"
  }
}

Configure retention of assets generated by asynchronous query api or scheduled reports

By default the MPF is configured to retain reports generated by the asynchronous query api and scheduled reports generated by CFEngine Enterprise. This can be configured by setting vars.purge_scheduled_reports_older_than_days via augments.

{
  "vars": {
    "purge_scheduled_reports_older_than_days": "30"
  }
}

Adjust the maximum amount of client side report data to retain (CFEngine Enterprise)

Enterprise agents cache detailed information about each agent run locally. The data is purged when the data is reported to a hub. If the volume of data exceeds def.max_client_history_size then the client will purge the local data in order to keep report collection from timing out.

The default 50M threshold can be configured using an [augments file][Augments], for example:

{
  "vars": {
    "max_client_history_size": "5M"
  }
}

Enterprise hub pull collection schedule

By default Enterprise hubs initiate pull collection once every 5 minutes. This can be overridden in the MPF by defining def.control_hub_hub_schedule as a list of classes that should trigger collection when defined.

Here we set the schedule to initiate pull collection once every 30 minutes via augments.

{
  "vars": {
    "control_hub_hub_schedule": [ "Min00", "Min30" ]
  }
}

History:

• MPF override introduced in 3.13.0, 3.12.2

Configure maximum age in hours of old reports for cf-hub to collect

By default cf-hub instructs clients to expire reports older than 6 hours in order to prevent a build up of reports that could cause a condition where the client is never able to send all reports within the collection window.

You can adjust this time by setting vars.control_hub_client_history_timeout

For example:

{
  "vars": {
    "control_hub_client_history_timeout": "72"
  }
}

History:

• MPF override introduced in 3.13.0, 3.12.2

Exclude hosts from hub initiated report collection

You may want to exclude some hosts like community agents, hosts behind NAT, and hosts using client initiated reporting from hub initiated report collection. To exclude hosts from hub initiated report collection define def.control_hub_exclude_hosts in an [augments file][Augments].

For example to completely disable hub initiated report collection:

{
  "vars": {
    "control_hub_exclude_hosts": [ "0.0.0.0/0" ]
  }
}

History:

• MPF override introduced in 3.13.0, 3.12.2

Change port used for enterprise report collection

By default cf-hub collects reports by connecting to port 5308. You can change this default by setting vars.control_hub_port in augments.

For example:

{
  "vars": {
    "control_hub_port": "8035"
  }
}

History:

• Added in 3.13.0, 3.12.2

Change hub to client connection timeout

By default, cf-hub times out a connection after 30 seconds. This can be configured in augments.

For example:

{
  "vars": {
    "control_hub_query_timeout": "10"
  }
}

Note:

• A value of "0" will cause the default to be used.

History:

• Added in 3.15.0

Enable client initiated reporting

In the default configuration for Enterprise report collection the hub periodically polls agents that are bootstrapped to collect reports. Sometimes it may be desirable or necessary for the client to initiate report collection.

To enable client initiated reporting define the class client_initiated_reporting_enabled. You may also want to configure the report interval (how frequently an agent will try to report it's data to the hub) by default it is set to 5. The reporting interval def.control_server_call_collect_interval and the class can be defined in an [augments file][Augments].

For example:

{
  "classes" {
    "client_initiated_reporting_enabled": [ "any" ]
  },
  "vars": {
    "control_server_call_collect_interval": "1",
  }
}

Configure client initiated reporting timeout

By default cf-serverd holds an open connection for client initiated for 30 seconds. In some environments this value may need to be increased in order for report collection to finish. Once the connection has been open for longer than the specified seconds it is closed.

The window of time can be controled by setting def.control_server_collect_window.

For example, enable client initated reporting for all hosts with a 10 minute interval and hold the connection open for 90 seconds.

{
  "classes" {
    "client_initiated_reporting_enabled": [ "any" ]
  },
  "vars": {
    "control_server_collect_window": "90",
    "control_server_call_collect_interval": "10"
  }
}

History:

• Added in 3.10.6, 3.12.2, 3.13.1

Configure MPF to automatically restart components on relevant data change

While the agent itsef will reload its config upon notice of policy change this bundle specifically handles changes to variables used in the MPF which may come from external data sources which are unknown to the components themselves.

Currently only cf-serverd, cf-monitord, and cf-hub are handled. cf-execd is NOT automatically restarted.

To enable this functionality define the class mpf_augments_control_enabled

{
  "classes":{
      "mpf_augments_control_enabled": [ "any" ]
  }
}

Notes: In order for custom ACLs to leverage augments and support data based restart you should use variables prefixed with control_server_.

For example changes to vars.control_server_my_access_rules when mpf_augments_control_enabled is defined will result in cf-serverd restarting.

{
  "classes": {
      "mpf_augments_control_enabled": [ "any" ]
  },
  "vars": {
    "control_server_my_access_rules" : {
      "/var/repo/": {
        "admit": "def.acl"
      }
    }
  }
}

History: Added 3.11.0

Configure maxconnections for cf-serverd

maxconnections in body server control configures the maximum number of connections allowed by cf-serverd. Recommended to be set greater than the number of hosts bootstrapped.

This can be configured via [augments][Augments]:

{
  "vars":{
      "control_server_maxconnections": "1000"
  }
}

History: Added 3.11.0

Configure networks allowed to make collect_calls (client initiated reporting)

By default the hub allows collect calls (client initiated reporting) from the networks defined in def.acl To configure which networks are allowed to initiate report collection define def.mpf_access_rules_collect_calls_admit_ips.

For example to allow client initiated reporting for hosts coming from 24.124.0.0/16:

{
  "vars": {
    "mpf_access_rules_collect_calls_admit_ips": [ "24.124.0.0/16" ]
  }
}

See Also: Generic acl

Configure Enterprise Measurement/Monitoring Collection

Metrics recorded by measurement promises in cf-monitord are only collected by default for policy servers. In order to collect metrics for non policy servers simply define default_data_select_host_monitoring_include via in an [augments file][Augments].

For example to collect all measurements for remote agents and only cpu and memory related probes on policy servers:

{
  "vars": {
    "default_data_select_host_monitoring_include": [ ".*" ],
    "default_data_select_policy_hub_monitoring_include": [ "mem_.*", "cpu.*" ]
  }
}

History:

• Added in 3.10.2, 3.11.0

Configure Enterprise Mission Portal Docroot

Primarily for developer conveniance, this setting allows you to easily disable the enforcement that the webapp consists of the packaged files in the docroot used for Mission Portal.

{
  "classes": {
    "mpf_disable_mission_portal_docroot_sync_from_share_gui": [ "any" ]
  }
}

Bundlesequence

Classification bundles before autorun

You can specify a list of bundles which should be run before autorun policies (if enabled).

{
  "vars":{
    "control_common_bundlesequence_classification": [ "classification_one", "classification_two" ]
  },

  "inputs": [ "services/my_classificaton.cf" ]
}

History:

• Added in CFEngine 3.18.0

Append to the main bundlesequence

You can specify bundles which should be run at the end of the default bundlesequence by defining control_common_bundlesequence_end in the vars of an [augments file][Augments].

For example:

{
  "vars":{
    "control_common_bundlesequence_end": [ "mybundle1", "mybundle2" ]
  },

  "inputs": [ "services/mybundles.cf" ]
}

Notes:

• The order in which bundles are actuates is not guaranteed.
• The agent will error if a named bundle is not part of inputs.

History: Added in 3.10.0

Configure abortclasses via augments

Configure a list of regular expressions that result in cf-agent terminating itself upon definition of a matching class.

For example, abort execution if any class starting with error_ or abort_ is defined:

{
  "vars":{
    "control_agent_abortclasses": [ "error.*", "abort.**" ]
  }
}

History: Added in 3.15.0, 3.12.3

Configure abortbundleclasses via augments

Configure a list of regular expressions that match classes which if defined lead to termination of current bundle.

If no list is defined, then a default of abortbundle is used.

For example, abort execution if any class starting with bundle_error_ or bundle_abort_ is defined:

{
  "vars":{
    "control_agent_abortbundleclasses": [ "bundle_error.*", "bundle_abort.**" ]
  }
}

History: Added in 3.15.0, 3.12.3

Configure files_single_copy via augments

Specify a list of regular expressions that when matched will prevent the agent from performing subsequent copy operations on the same promiser.

For example, to only allow any file to be copied a single time:

{
  "vars":{
    "control_agent_files_single_copy": [ ".*" ]
  }

}

History: Added in 3.11.0, 3.10.2

Disable automatic policy hub detection

During bootstrap, if the executing host finds the IP address of the target on itself it automatically classifies the host as a policy server by ensuring the $(sys.statedir)/am_policy_hub file exists. When this file exists, the am_policy_hub and policy_server classes are defined. To help avoid accidental declassification, the MPF contains policy to regularly check if the host is bootstrapped to an IP found on itself, and if so, to ensure the proper state file exists.

To disable this check, define mpf_auto_am_policy_hub_state_disabled.

For example, to define this class via augments, place the following in your def.json.

{
  "classes":{
    "mpf_auto_am_policy_hub_state_disabled": [ "any" ]
  }

}

History: Added in 3.15.0, 3.12.3, 3.10.7

Configure default repository for file backups

By default the agent creates a backup of a file before it is edited in the same directory as the edited file. Defining the mpf_control_agent_default_repository class will cause these backups to be placed in $(sys.workdir)/backups. Customize the backup directory by setting def.control_agent_default_backup.

For example:

{
  "classes": {
    "mpf_control_agent_default_repository": [ "any" ]
  },

  "vars": {
    "control_agent_default_repository": "/var/cfengine/edit_backups"
  }
}

History: Added in 3.10.1

Configure periodic package inventory refresh interval

Note that there are currently two implementations of packages promises, package modules and package methods. Each maintain their own cache of packages installed and updates available.

For package modules

CFEngine refreshes software inventory when it makes changes via packages promises. Additionally, by default, CFEngine refreshes it's internal cache of packages installed (during each agent run) and package updates that are available (once a day) according to the default package manager in order to pick up changes made outside packages promises.

{
  "vars": {
    "package_module_query_installed_ifelapsed": "5",
    "package_module_query_updates_ifelapsed": "60"
  }
}

WARNING: Be ware of setting package_module_query_update_ifelapsed too low, especially with public repositories or you may be banned for abuse.

See also: packagesmatching(), packageupdatesmatching()

History:

• Added in 3.15.0, 3.12.3
• 3.17.0 decreased package_module_query_installed_ifelapsed from 60 to 0

For package methods

CFEngine refreshes it's cache of information about packages installed and updates available when it evaluates packages promises if the cache has not been updated in the number of minutes stored in package_list_update_ifelapsed of the package method in use. Many package methods in the standard library use the value of default:common_knowledge.list_updates_ifelapsed for this value which can be customized via Augments.

{
  "variables": {
    "default:common_knowledge.list_update_ifelapsed": {
      "value": "0"
    }
  }
}

Notes:

• Unlike many variables that can be customized via Augments this variable is not in the default:def bundle scope. Customizing it requires CFEngine 3.18.0 or newer which support targeting any namespace or variable.

See also:

• [package methods][lib/packages.cf]: pip, npm, npm_g, brew, apt, apt_get, apt_get_permissive, apt_get_release, dpkg_version, rpm_version , yum, yum_rpm, yum_rpm_permissive, yum_rpm_enable_repo , yum_group, rpm_filebased, ips, smartos, opencsw, emerge, pacman, zypper, generic

• [package bundles][lib/packages.cf]: package_latest, package_specific_present, package_specific_absent, package_specific_latest, package_specific

History:

• Added in 3.22.0

Enable logging of Enterprise License utilization

If the class enable_log_cfengine_enterprise_license_utilization is defined on an enterprise hub license utilization will be logged by the hub in $(sys.workdir)/log/license_utilization.log

Example enabling the class from an [augments file][Augments]:

{
  "classes": {
    "enable_log_cfengine_enterprise_license_utilization": [ "enterprise_edition" ]
  }
}

History: Added in 3.11, 3.10.2

Enable external watchdog

Note: This feature is not enabled by default.

If the class cfe_internal_core_watchdog_enabled is defined, the feature is enabled and the watchdog will be active. If the class cfe_internal_core_watchdog_disabled is defined, the feature is disabled and the watchdog will not be active.

{
    "classes": {
        "cfe_internal_core_watchdog_enabled": [ "aix::" ],
        "cfe_internal_core_watchdog_disabled": [ "!cfe_internal_core_watchdog_enabled::" ]
        }
}

See Also: [Watchdog documentation][cfe_internal/core/watchdog]

Environment Variables

Environment variables that should be inherited by child commands can be set using def.control_agent_environment_vars_default. The policy defaults are overridden if this is defined. This can be useful if you want to modify the default environment variables that are set.

For example:

{
  "vars": {
    "control_agent_environment_vars_default":
      [ "DEBIAN_FRONTEND=noninteractive",
        "XPG_SUS_ENV=ON" ]
  }
}

The environment variables can also be extended by defining def.control_agent_environment_vars_extra. The extra environment variables defined here are combined with the defaults (if they exist).

{
  "vars": {
    "control_agent_environment_vars_extra": [ "XPG_SUS_ENV=ON" ]
  }
}

Notes:

• Simple augments as shown above apply to all hosts. Consider using the [augments key][Augments#augments] or [host specific data][Augments#host_specific.json] if you want to set environment variables differently across different sets of hosts. The value set via Augments takes precedence over policy defaults, so be sure to take that into account when configuring.

History:

• Introduced in 3.20.0, 3.18.2

Modules

Modules executed by the usemodule() function are expected to be found in $(sys.workdir)/modules the modules are distributed to all remote agents by in the default policy.

Templates

For convenience the templates shortcut is provided and by default the path is set to $(sys.workdir/templates) unless $(def.template_dir) is overridden via [augments][Augments].

• NOTE: The templates directory is not currently managed by default policy. Unlike modules templates are not distributed to all hosts by default.

Copy a template from the templates directory:

  files:

    "$(def.dir_templates)/mytemplate.mustache" -> { "myservice" }
      copy_from => remote_dcp("templates/mytemplate.mustache", $(sys.policy_server) ),
      comment => "mytemplate is necessary in order to render myservice configuration file.";

Override the path for $(def.dir_templates) by setting vars.dir_templates in the [augments file][Augments] (def.json):

{
    "vars": {
        "dir_templates": "/var/cfengine/mytemplates"
        }
}

Note: When overriding the templates directory a change to the [augments][Augments] alone will not cause cf-serverd to reload its configuration and update the access control lists as necessary. cf-serverd will only automatically reload its config when it notices a change in policy.

History: Added in 3.11.

Federated Reporting

Enable Federated Reporting Distributed Cleanup

Hosts that report to multiple feeders result in duplicate entries and other issues. Distributed cleanup helps to deal with this condition.

To enable this functionality define the class default:cfengine_mp_fr_enable_distributed_cleanup on the superhub.

For example, to define this class via Augments:

{
  "classes": {
    "cfengine_mp_fr_enable_distributed_cleanup": [ "any::" ]
  }
}

History:

• Added in CFEngine 3.19.0, 3.18.1

Configure SSL Certificate Directory for Federated Reporting Distributed Cleanup

When custom certificates are in use distributed cleanup needs to know where to find them. To configure the path where certificates are found define default:def.DISTRIBUTED_CLEANUP_SSL_CERT_DIR, for example:

{
  "vars": {
    "DISTRIBUTED_CLEANUP_SSL_CERT_DIR": "/path/to/my/cert/dir"
  }
}

History:

• Added in CFEngine 3.20.0, 3.18.2

PostgreSQL Configuration

It's not uncommon to need to configure some PostgreSQL settings differently for Federated Reporting. The settings that are exposed as tunables which can be set via augments are listed here. These do not comprise all settings that may need adjusted, only those that are most commonly adjusted.

Note: When setting parameters for the PostgreSQL configuration file various units can be used. Valid memory units are B (bytes), kB (kilobytes), MB (megabytes), GB (gigabytes), and TB (terabytes). The multiplier for memory units is 1024, not 1000. Valid time units are us (microseconds), ms (milliseconds), s (seconds), min (minutes), h (hours), and d (days).

shared_buffers

Shared buffers are the amount of memory the database server uses for shared memory buffers. Settings significantly higher than the minimum are usually needed for good performance.

The value should be set to 15% to 25% of the machine's total RAM. For example: if your machine's RAM size is 32 GB, then the recommended value for shared_buffers is 8 GB.

To adjust this set cfengine_enterprise_federation:postgres_config.shared_buffers via Augments.

For example:

{
  "variables": {
    "cfengine_enterprise_federation:postgres_config.shared_buffers": "2560MB"
  }
}

History:

• Added in 3.20.0, 3.18.2

max_locks_per_transaction

The max_locks_per_transaction value indicates the number of database objects that can be locked simultaneously. When Federated Reporting is enabled, the MPF default is 4000.

{
  "variables": {
    "cfengine_enterprise_federation:postgres_config.max_locks_per_transaction": "4100"
  }
}

History:

• Added in 3.20.0, 3.18.2

log_lock_waits

Controls whether a log message is produced when a session waits longer than deadlock_timeout to acquire a lock. This is useful in determining if lock waits are causing poor performance. When Federated Reporting is enabled, the MPF default is on.

{
  "variables": {
    "cfengine_enterprise_federation:postgres_config.log_lock_waits": "off"
  }
}

History:

• Added in 3.20.0, 3.18.2

max_wal_size

Sets the WAL size that triggers a checkpoint.

Maximum size to let the WAL grow during automatic checkpoints. This is a soft limit; WAL size can exceed max_wal_size under special circumstances, such as heavy load, a failing archive_command, or a high wal_keep_size setting. If this value is specified without units, it is taken as megabytes. The default is 1 GB (1024MB). Increasing this parameter can increase the amount of time needed for crash recovery.

{
  "variables": {
    "cfengine_enterprise_federation:postgres_config.max_wal_size": "20G"
  }
}

History:

• Added in 3.20.0, 3.18.2

checkpoint_timeout

Sets the maximum time between automatic WAL checkpoints.

Maximum time between automatic WAL checkpoints. If this value is specified without units, it is taken as seconds. The valid range is between 30 seconds and one day. The default is five minutes (5min). Increasing this parameter can increase the amount of time needed for crash recovery.

{
  "variables": {
    "cfengine_enterprise_federation:postgres_config.checkpoint_timeout": "30min"
  }
}

History:

• Added in 3.20.0, 3.18.2

Recommendations

The MPF includes policy that inspects the system and makes recommendations about the configuration of the system. When cfengine_recommendations_enabled is defined bundles tagged cfengine_recommendation are executed in lexical order. cfengine_recommendations_enabled is defined by default when cfengine_recommendations_disabled is not defined.

To disable cfengine recommendations define cfengine_recommendations_disabled.

This snippet disables recommendations via augments.

{
    "classes": [
        "cfengine_recommendations_disabled"
    ]
}

History:

• Recommendations added in 3.12.1