Adoptable Cookbooks List

Looking for a cookbook to adopt? You can now see a list of cookbooks available for adoption!
List of Adoptable Cookbooks

Supermarket Belongs to the Community

Supermarket belongs to the community. While Chef has the responsibility to keep it running and be stewards of its functionality, what it does and how it works is driven by the community. The chef/supermarket repository will continue to be where development of the Supermarket application takes place. Come be part of shaping the direction of Supermarket by opening issues and pull requests or by joining us on the Chef Mailing List.

Select Badges

Select Supported Platforms


opennms (54) Versions 2.0.0

Installs and Configures opennms and provides many useful LWRPs.

cookbook 'opennms', '= 2.0.0'
cookbook 'opennms', '= 2.0.0', :supermarket
knife cookbook site install opennms
knife cookbook site download opennms


A Chef cookbook to manage the installation and configuration of OpenNMS. Current version of templates are based on OpenNMS release 15.0.2


Things generally work. Use the latest tag. Tags happen when a decent chunk of new functionality is written, their example recipes converge (named example_*) and OpenNMS starts without errors. Notably missing are unit and more formal integration tests.


  • Chef 11.x+
  • CentOS 6.x. Debian/Ubuntu support shouldn't be too hard to do - if anyone wants to head that up let me know.
  • Either use Berkshelf to satisfy dependencies or manually acquire the following cookbooks:
    • yum
    • hostsfile
  • Unless you like really old JVMs and a poorly tuned PostgreSQL or just really want to install java and postgresql yourself you will also want these cookbooks:
    • java
    • postgresql

See Usage for more details.


Running the default recipe will install OpenNMS 15.0.2 (or a custom version using the attribute node[:opennms][:version]) on CentOS 6.x from the official repo with the default configuration. It will also execute '$ONMS_HOME/bin/runjava -s if $ONMS_HOME/etc/java.conf is not present and $ONMS_HOME/bin/install -dis if $ONMS_HOME/etc/configured is not present.

There are two primary ways to use this cookbook: as an application cookbook or library cookbook. If you simply want to tweak a few settings to the default OpenNMS configuration, you can use the default recipe of this cookbook directly and modify node attributes to suit your needs. There are also a plethora of LWRPs that you can use to do more in depth customizations. If you go that route I recommend starting with the notemplates recipe and then using those LWRPs (and maybe a few of the templates in this cookbook) to define your run list. If your node's run list contains both the template and a resource that manages the same file you'll end up with a lot of churn during the chef client run, which is a waste of time and will probably cause unnecessary restarts of OpenNMS.

Template resources for daemons that support configuration changes without a restart will automatically send the proper event to activate changes. Add notifies to your resource for similar funcationality when using the LWRPs from this cookbook. See the example recipe for each LWRP for details.

You probably also want to check out the community java ( and postgresql ( cookbooks. Here's an example of each:


You probably want Java 8 because Java 7 is EOL. But, the community java cookbook doesn't support Java 8 as an RPM because reasons. I don't like installing things outside of package management if it can be avoided, so if you host the RPM yourself and use dschlenk's fork you can get around that limitation. Here's an example of how:

Download the appropriate RPM(s) from Oracle and make a yum repo available to your nodes.

Use the link above by either cloning the repo and uploading to your Chef server or using Berkshelf or another cookbook management tool that can talk to git repos.

Set the following attributes:

node['java']['oracle']['accept_oracle_download_terms'] = true
node['java']['install_flavor'] = 'oracle_rpm'
node['java']['oracle_rpm']['type'] = 'jdk'
node['java']['oracle_rpm']['package_name'] = 'jdk1.8.0_40'
node['java']['alternatives_priority'] = 180040
node['java']['jdk_version'] = 8
node['java']['set_etc_environment'] = true
node['java']['oracle']['jce']['enabled'] = true

Add a yum_repository resource to your run list, like so:

yum_repository 'oracle-java' do
  description 'mirror of oracle java RPM packages'
  baseurl 'URL_TO_YOUR_YUM_REPO'
  gpgcheck false
  action :create

Then add the 'java::default' recipe to your run list.


Include the client, server, contrib, config_initdb, config_pgtune recipes (in that order) in your run list. Then use these attributes for a fairly well tuned config:

node[:postgresql][:pg_hba] = { :addr => '' :user => 'all', :type => 'local', :method => 'trust', :db => 'all' }
node[:postgresql][:pg_hba] = { :addr => '', :user => 'all', :type => 'host', :method => 'trust', :db => 'all'}
node[:postgresql][:pg_hba] = { :addr => '::1/128', :user => 'all', :type => 'host', :method => 'trust', :db => 'all' }
node[:postgresql][:config][:checkpoint_timeout] = '15min'
node[:postgresql][:config][:autovacuum] = 'on'
node[:postgresql][:config][:track_activities] = 'on'
node[:postgresql][:config][:track_counts]  = 'on'
node[:postgresql][:config][:shared_preload_libraries] = 'pg_stat_statements'
node[:postgresql][:config][:vacuum_cost_delay] = 50
node[:postgresql][:client][:packages] = ["postgresql", "postgresql-contrib", "postgresql-devel"]
node[:postgresql][:server][:packages] = ["postgresql-server"]

Or, if you want to use a version of Postgresql from this decade, use these attributes:

node['postgresql']['enable_pgdg_yum'] = true
node['postgresql']['version'] = '9.3'
node['postgresql']['dir'] = '/var/lib/pgsql/9.3/data'
node['postgresql']['config']['data_directory'] = '/var/lib/pgsql/9.3/data'
node['postgresql']['config']['autovacuum'] = 'on'
node['postgresql']['config']['checkpoint_timeout'] = '15min'
node['postgresql']['config']['shared_preload_libraries'] =
node['postgresql']['config']['track_activities'] = 'on'
node['postgresql']['config']['track_counts'] = 'on'
node['postgresql']['config']['vacuum_cost_delay'] = 50
node['postgresql']['pg_hba'] = [{ addr: '', db: 'all', method: 'trust',
                                     type: 'local', user: 'all' },
                                   { addr: '', db: 'all',
                                     method: 'trust', type: 'host',
                                     user: 'all' },
                                   { addr: '::1/128', db: 'all',
                                     method: 'trust', type: 'host',
                                     user: 'all' }]
node['postgresql']['client']['packages'] = ['postgresql93',
node['postgresql']['server']['packages'] = ['postgresql93-server']
node['postgresql']['contrib']['packages'] = ['postgresql93-contrib']
node['postgresql']['server']['service_name'] = 'postgresql-9.3'
node['postgresql']['contrib']['extensions'] = ['pageinspect',

There are also a couple OpenNMS attributes you'll probably want to override at a minimum:


node[:opennms][:conf][:start_timeout] = 20
node[:opennms][:conf][:heap_size] = 1024
node[:opennms][:conf][:addl_mgr_opts] = '-XX:+UseConcMarkSweepGC -XX:MaxMetaspaceSize=512m'

The last part (-XX:MaxMetaspaceSize=512m) assumes Java 8 and will cause OpenNMS to not start if you use Java 7 or earlier. It's the modern equivalent of setting -XX:MaxPermSize=512m, which is even more important in Java 8 because things that used to be in PermGen are now in Heap and it grows unbounded unless you tell it not to.


Starting with version 2.0.0 there is now experimental support for handling upgrades automatically. Use at your own risk. It is disabled by default. To enable, set node['opennms']['upgrade'] to true. If this sounds like something you want to do, review the upgrade recipe. It roughly translates to:

  • New RPM is installed.
  • Are there any files named *.rpmnew in $ONMS_HOME/etc or $ONMS_HOME/jetty-webapp? If so, overwrite the existing files with them.
  • Are there any files named *.rpmsave in $ONMS_HOME/etc or $ONMS_HOME/jetty-webapp? If so, remove them.

rpmsave files happen when there's a config file that you have changed that was replaced with the new version because not replacing it would prevent OpenNMS from working properly. But since we're using Chef, we don't care about the old version as any changes we made to it previously will be redone with the appropriate LWRP(s) or templates later in the converge. Since OpenNMS won't start with these files in place we just remove them.

Similarly, rpmnew files are created when a newer version of a file exists, but it doesn't contain breaking changes. Just like rpmsave files, OpenNMS won't start with these files present, and the rest of the converge will make the changes we want anyway, so we just overwrite the old file with the rpmnew version.

There is also an attribute named node['opennms']['allow_downgrade'] which if set to true will let you install an older version of OpenNMS than is currently installed. Leave this alone unless you know what you're doing - sometimes yum gets confused with versions when working with snapshot builds and this attribute helps work around that.


  • opennms::default Installs and configures OpenNMS with the standard configuration modified with any node attribute values changed from their defaults.
  • opennms::notemplates Everything default does except minimal templates are used - etc/opennms.conf, etc/ and etc/jetty.xml. Use this recipe if you intend to use any of the LWRPs in this cookbook.
  • opennms::nsclient installs the optional nsclient data collection plugin and uses the template for etc/nsclient-datacollection-config.xml.
  • opennms::xml installs the optional xml data collection plugin and uses the template for etc/xml-datacollection-config.xml.


As a general rule these LWRPs support a single action: create and many of them behave more like create_if_missing does in other cookbooks. In other words, updating is generally not supported. This may change in future releases.

Also, there are example recipes in the cookbook for most every LWRP named opennms::example_<LWRP_NAME>. Eventually these will become tests.

The list of implemented LWRPs is as follows:

Users, Groups and Roles

  • opennms_user: add a user. Uses the REST API.
  • opennms_group: add a group and populate it with users. You can even set the default SVG map and duty schedules.
  • opennms_role: add a role.
  • opennms_role_schedule: Add schedules to a role. See an example for this and the role LWRP in recipe opennms::example_role.


  • opennms_disco_specific: add a specific IP to be discovered.
  • opennms_disco_range: add a include or exclude range discovery.
  • opennms_disco_url: add a include-url to discovery and if it's a file deploy it where specified.

Provisioning Requisitions

These LWRPs use a cookbook library named Provision that I wrote to perform the work using the OpenNMS REST interface. As such, OpenNMS has to be running for the resources to converge. Also you'll notice that I used the term 'import' rather than the correct term 'requisition'. I can type 'import' a lot faster than 'requisition'. ;)

  • opennms_foreign_source: create a new foreign source optionally defining a scan interval (defaults to '1d').
  • opennms_service_detector: add a service detector to a foreign source.
  • opennms_policy: add a policy to a foreign source.
  • opennms_import: Defines a requisition for a foreign source. This and all import* LWRPs include an option to synchronize the requisition - sync_import.
  • opennms_import_node: Add a node to a requisition including categories (array of strings) and assets (key/value hash pairs).
  • opennms_import_node_interface: Add an interface to a node in a requisition.
  • opennms_import_node_interface_service: Add a service to an interface on a node in a requisition.


  • opennms_eventconf: adds an event-file element to events in etc/eventconf.xml.
  • opennms_event: adds an event element to events in target eventconf file file. Not all elements from the eventconf schema are implemented, but the ones that seem to actually exist in the wild are. See resource for details and recipes example_event and example_threshold for example usage.
  • opennms_send_event: creates an actual instance of an event using the script in $ONMS_HOME/bin. Used by the send_events recipe, which is included by default and notemplates recipes to cause config file reloads to take place when template resources make changes or an LWRP sends a notification.


  • opennms_notification_command: Create a new command in notificationCommands.xml.
  • opennms_destination_path: creates a destination path element in destinationPaths.xml. Requires at a minimum a single target which can be defined with the following LWRP.
  • opennms_target: Add a target or escalate target to a destination path (defined either in the default config or with the above LWRP).
  • opennms_notification: Create notification elements in notifications.xml.
  • opennms_notifd_autoack: Create a auto acknowledge up/down pair in notifd-configuration.xml

Node Service Credential Configuration

These LWRPs allow you to define the credentials necessary to connect to services on monitored nodes. These are some of the few that currently implement updating and deleting. Action :create will update if changes are detected but :create_if_missing will do nothing. To determine if a resource needs to be updated or deleted, existance is determined by all definition element attributes being equal (so all resource attributes except ranges, specifics, ip_matches and position).

If an update occurs, the values contained in the new resource will be used. Note that all range, specific and ip-match elements that exist currently in the definition will be removed before the new elements are added.

Currently implemented are:

  • opennms_snmp_config_definition: add a definition element to snmp-config.xml.
  • opennms_wmi_config_definition: add a definition element to wmi-config.xml.


  • opennms_poller_package: add a package to etc/poller-configuration.xml. Note that an instance of this resource without use of an accompanying opennms_poller_service resource will result in a failure to start opennms.
  • opennms_poller_service: add a service to poller package named poller_name. See opennms::example_poller for example usage of this and the opennms_poller_package resource.

Data Collection

  • opennms_resource_type: adds a resourceType definition to a file in etc/datacollection and an include-collection element to the default snmp-collection. This LWRP supports a very limited form of updating - if the resource type already exists but isn't included in the default snmp-collection, an include-collection element will be added. The definition of the resource type won't be updated, however.
  • opennms_system_def: add or remove pre-existing groups (/datacollection-group/group[@name]) to or from pre-existing systemDefs (/datacollection-group/systemDef/collect/includeGroup[text()]) in $ONMS_HOME/etc/datacollection/*.xml.
  • opennms_snmp_collection: adds an snmp-collection element to etc/datacollection-config.xml.
  • opennms_xml_collection: adds an xml-collection element to etc/xml-datacollection-config.xml.
  • opennms_wmi_collection: adds a wmi-collection element to etc/wmi-datacollection-config.xml.
  • opennms_jdbc_collection: adds a jdbc-collection element to etc/jdbc-datacollection-config.xml.
  • opennms_collection_package: adds a package element to etc/collectd-configuration.xml.
  • opennms_snmp_collection_service: adds a service element to a package in etc/collectd-configuration.xml.
  • opennms_xml_collection_service: adds a service element to a package in etc/collectd-configuration.xml.
  • opennms_wmi_collection_service: adds a service element to a package in etc/collectd-configuration.xml.
  • opennms_jdbc_collection_service: adds a service element to a package in etc/collectd-configuration.xml.
  • opennms_snmp_collection_group: adds an include-collection element to an snmp-collection in etc/datacollection-config.xml and drops off the specified cookbook file into etc/datacollection.
  • opennms_jdbc_query: adds a query element to a jdbc-collection in etc/jdbc-datacollection-config.xml.
  • opennms_xml_source: adds a xml-source element to a xml-collection in etc/xml-datacollection-config.xml.
  • opennms_xml_group: adds a xml-source element to a xml-source in etc/xml-datacollection-config.xml.

Statistics Reports

See opennms::example_statsd for example usage of these LWRPs.

  • opennms_statsd_package: create a new package optionally with a filter in statsd-configuration.xml.
  • opennms_statsd_report: add a report to a package in statsd-configuration.xml.


  • opennms_collection_graph_file: Add a cookbook file containing graph definitions (perhaps generated by the mib compiler) to $ONMS_HOME/etc/
  • opennms_collection_graph: Add a new graph definition to the main (bad idea), new or an existing graph file.
  • opennms_response_graph: Add a response graph to $ONMS_HOME/etc/ Since there's a pretty well defined pattern to these, you can define these with just the name of the data source and it'll create a graph with min, max and average response times.


See examples for all of these LWRPs are in a single recipe, example_threshold.

  • opennms_threshd_package: Create a new package in threshd-configuration.xml.
  • opennms_threshold_group: Create a new threshold group in thresholds.xml.
  • opennms_threshold: Create a new threshold in the specified group in thresholds.xml.
  • opennms_expression: Create a new expression threshold in the specified group in thresholds.xml.

Web UI

There are a couple LWRPs for managing the Web UI. All of these support updating.

  • opennms_avail_category: Define categories for use in the Availability box on the main page (and the Summary dashlet in Ops Board).
  • opennms_avail_view: Define the list of categories in each view sections displayed in the Availability box on the main page (and the Summary dashlet in Ops Board).
  • opennms_wallboard: Create a wallboard.
  • opennms_dashlet: Add a dashlet to a wallboard.
  • opennms_surveillance_view: Manage surveillance views used in the legacy dashboard, Ops Board, and optionally on the front page. Note: does not verify that the categories you reference exist, because there's no ReST interface (yet).

Template Overview

Most configuration files are templated and can be overridden with environment, role, or node attributes. See the default attributes file for a list of configuration items that can be changed in this manner, or keep reading for a brief overview of each template available. Default attribute values set to nil mean that the file's default value is commented out in the original file and will remain so unless set to a non-nil value.

Access Point Monitor (etc/access-point-monitor-configuration.xml)

You can change threads and pscan_interval or disable either of the default types by setting aruba_enable or moto_enable to false in node['opennms']['apm'].


If you want to change the logo or default interval, count, hour or minute you can do so for either the calandar or classic report like so: { "opennms": { "db_reports": { "avail": { "cal": { "logo": "/path/to/an/other_logo.png" }, "classic": { "count": 7 } } } } }


Note: the existance of this template does not imply that you should still be using capsd! To enable capsd you need to change the handler for newSuspect events in opennms.conf by setting node['opennms']['properties']['provisioning']['enable_discovery'] to false.

Example: change the timeout and retries for the ICMP scanner, do:

     "opennms": {
       "capsd": {
         "icmp": {
           "timeout": "5000",
           "retry": "3"


Default categories can be modified by doing things like

     "opennms": {
       "categories": {
         "web": {
           "services": [

That'll leave the defaults for everything except overwrite the list of services in the web category. The names of the categories are: 'overall', 'interfaces', 'email', 'web', 'jmx', 'dns', 'db', 'other', 'inet'. The defaults are nil, which leaves the defaults as is. Override with false to disable.


Disable one of the three default charts by setting severity_enabled, outages_enable or inventory_enable to false in node['opennms']['chart'].


Use the node['opennms']['collectd']['threads'] attribute to change the number of threads (duh). There are also attributes for each default package. As of 1.12.5 those are: vmware3, vmware4, vmware5, example1. To modify one of those, for example, to change the IPv4 include range in the example1 package, you would do: { "opennms": { "collectd": { "example1": { "ipv4_range": { "begin": "", "end": "" } } } } } That leaves most of the example1 package as default. Set a package's enabled attribute to false if you want to completely remove that package. You can also do that for specific services in that package. See the template for more options.


You can override some settings like: { "opennms": { "datacollection": { "default": { "snmpStorageFlag": "all" } } } Or maybe you don't have any Dell gear: { "opennms": { "datacollection": { "default": { "dell": false } } } }

You can also remove one of the default snmp-collections, or change the step and RRA definitions.


Attributes are available in node['opennms']['discovery'] to change global settings: * threads (threads) * packets-per-second (pps) * initial-sleep-time (init_sleep_ms) * restart-sleep-time (restart_sleep_ms) * retries (retries) * timeout (timeout)


Attributes are available in node['opennms']['eventd'] to change global settings: * TCPAddress (tcp_address) * TCPPort (tcp_port) * UDPAddress (udp_address) * UDPPort (udp_port) * receivers (receivers) * socketSoTimeoutRequired (sock_so_timeout_req to true or false) * socketSoTimeoutPeriod (socket_so_timeout_period)


Attributes are available in node['opennms']['events_archiver'] to change global settings: * archiveAge (age) * separator (separator)


This file controls how OpenNMS sends email. This is not where you configure the mail monitor. Attributes available in node['opennms']['javamail_props']. They follow the config file but with ruby style because the kids hate camel case I guess. * org.opennms.core.utils.fromAddress (from_address) * org.opennms.core.utils.mailHost (mail_host) * ...and so on.


This is where you configure the mail monitor. Attributes available in node['opennms']['javamail_config'. Unlike most of the templates, you can change every attribute and element in the default sendmail and receivemail elements since the defaults are useful to no one. Here's a list of the defaults which you definitely need to override if you want a mail monitor to work:

default['opennms']['javamail_config']['default_read_config_name'] = "localhost"
default['opennms']['javamail_config']['default_send_config_name'] = "localhost"
default['opennms']['javamail_config']['default_read']['attempt_interval'] = 1000
default['opennms']['javamail_config']['default_read']['delete_all_mail']  = false
default['opennms']['javamail_config']['default_read']['mail_folder']      = "INBOX"
default['opennms']['javamail_config']['default_read']['debug']            = true
default['opennms']['javamail_config']['default_read']['properties']       = {'mail.pop3.apop.enable' => false, 'mail.pop3.rsetbeforequit' => false}
default['opennms']['javamail_config']['default_read']['host']             = ""
default['opennms']['javamail_config']['default_read']['port']             = 110
default['opennms']['javamail_config']['default_read']['ssl_enable']       = false
default['opennms']['javamail_config']['default_read']['start_tls']        = false
default['opennms']['javamail_config']['default_read']['transport']        = "pop3"
default['opennms']['javamail_config']['default_read']['user']             = "opennms"
default['opennms']['javamail_config']['default_read']['password']         = "opennms"
default['opennms']['javamail_config']['default_send']['attempt_interval']   = 3000
default['opennms']['javamail_config']['default_send']['use_authentication'] = false
default['opennms']['javamail_config']['default_send']['use_jmta']           = true
default['opennms']['javamail_config']['default_send']['debug']              = true
default['opennms']['javamail_config']['default_send']['host']               = ""
default['opennms']['javamail_config']['default_send']['port']               = 25
default['opennms']['javamail_config']['default_send']['char_set']           = "us-ascii"
default['opennms']['javamail_config']['default_send']['mailer']             = "smtpsend"
default['opennms']['javamail_config']['default_send']['content_type']       = "text/plain"
default['opennms']['javamail_config']['default_send']['encoding']           = "7-bit"
default['opennms']['javamail_config']['default_send']['quit_wait']          = true
default['opennms']['javamail_config']['default_send']['ssl_enable']         = false
default['opennms']['javamail_config']['default_send']['start_tls']          = false
default['opennms']['javamail_config']['default_send']['transport']          = "smtp"
default['opennms']['javamail_config']['default_send']['to']                 = "root@localhost"
default['opennms']['javamail_config']['default_send']['from']               = "root@[localhost]"
default['opennms']['javamail_config']['default_send']['subject']            = "OpenNMS Test Message"
default['opennms']['javamail_config']['default_send']['body']               = "This is an OpenNMS test message."
default['opennms']['javamail_config']['default_send']['user']               = "opennms"
default['opennms']['javamail_config']['default_send']['password']           = "opennms"

This is useful for something I'm sure, but I don't know what. See the template or default attributes file for hints.


Similar to other datacollection-config.xml files, you can change the RRD repository, step, RRA definitions and disable default collections and their queries.


Similar to other datacollection-config.xml files, you can change the RRD repository, step, RRA definitions and disable default collections and their mbeans. In the JBoss collection you can specify a JMS queue and/or topic to collect stats on. See the template and default attributes for details.


Attributes available in node['opennms']['linkd'] that allow you change global settings like: * threads * initial_sleep_time * snmp_poll_interval * discovery_link_interval

You can also turn off various kinds of detection, like for iproutes, set any of these to false to remove them from the file: * netscreen * cisco * darwin

Finally there's the package element at the end of the file that you can configure with these attributes: default['opennms']['linkd']['package'] = "example1" default['opennms']['linkd']['filter'] = "IPADDR != ''" default['opennms']['linkd']['range_begin'] = "" default['opennms']['linkd']['range_end'] = ""


This one is a little different. If you want to turn up logging for collectd, for instance, you'd set these override attributes: { "opennms": { "log4j": { "collectd": { "level": "DEBUG" } } } } But you could also change the appender class, the max file size, the max number of files (assuming the use of RollingFileAppender), the layout class and the layout conversion pattern (assuming the use of PatternLayout) using appender, max_file_size, max_backup_index, layout, and conversion_pattern.

The rtc username and password are populated from the values set in node['opennms']['properties']['rtc']['username'] and node['opennms']['properties']['rtc']['password']. TODO: Generate passwords during install! Other attributes available for configuration are: default['opennms']['magic_users']['admin_users'] = "admin" default['opennms']['magic_users']['ro_users'] = "" default['opennms']['magic_users']['dashboard_users'] = "" default['opennms']['magic_users']['provision_users'] = "" default['opennms']['magic_users']['remoting_users'] = "" default['opennms']['magic_users']['rest_users'] = "iphone"


Do you love maps but are a contrarian when it comes to color schemes? Have we got the template for you! I guess also useful for translating labels? Check out the default attributes for details on what you can change.


Join twitter and tell the public about your broken network! Set node['opennms']['microblog']['default_profile']['name'] to twitter or identica and then set ['opennms']['microblog']['default_profile']['authen_username'] and ['opennms']['microblog']['default_profile']['authen_password'] to use those services, or use a different service by setting node['opennms']['microblog']['default_profile']['service_url'] as well (assuming OpenNMS supports it). This only sets up the profile. You'll still need to define a destination path and set events and alarms to use it the normal way as described at until the notification and destination path LWRPs are written.


I couldn't figure out if this file is still used and/or necessary, but anyway the same attributes are used here and in provisiond-configuration.xml, so go check that one out.


You can set node['opennms']['modem'] to one of the predefined modems (mac2412, mac2414, macFA22, macFA24, macFA42, macFA44, acm0, acm1, acm2, acm3, acm4, acm5) or set node['opennms']['custom_modem'] to something like : { "opennms": { "custom_modem": { "name": "foobar", "port": "/dev/ttyFoobar", "model": "foobar37", "manufacturer": "Foo Bar" "baudrate": 57600 } } }


Is ignorance about your broken network in fact bliss? Shut off notifd by setting node['notifd']['status'] to "off" and find out. Don't know what match-all even means? Find out by setting node['opennms']['notifd']['match_all'] to false. (It controls whether only the first matching notification is used or not). You can also disable any of the default auto-acknowledge elements with node['notifd']['auto_ack']['service_unresponsive|service_lost|interface_down|widespread_outage'].


Turn off one of the default notification commands by setting one of the attributes in node['opennms']['notification_commands'] to false: * java_pager_email * java_email * xmpp_message * xmpp_group_message * irc_cat * call_work_phone * call_mobile_phone * call_home_phone * microblog_update * microblog_reply * microblog_dm


These attributes: * enabled * status * rule * destination_path * description * text_message * subject * numeric_message

can be overridden to alter any of these default notifications:

  • interface_down
  • node_down
  • node_lost_service
  • node_added
  • interface_deleted
  • high_threshold
  • low_threshold

in node['opennms']['notifications']. Stay tuned for a notification LWRP.


Similar to other datacollection-config.xml files, you can change the RRD repository, step, RRA definitions and disable default collections and wpms.


Disable packages, change IPv4 and IPv6 include ranges and RRD settings, remove or disable services or change their polling interval, retry, timeout or another service parameter. Like collectd and capsd, the possibilities here are pretty extensive so check out the default attributes and template for more info.


Same as default['opennms']['importer']['import_url'] = "file:/path/to/dump.xml" default['opennms']['importer']['schedule'] = "0 0 0 1 1 ? 2023" default['opennms']['importer']['threads'] = 8 default['opennms']['importer']['scan_threads'] = 10 default['opennms']['importer']['rescan_threads'] = 10 default['opennms']['importer']['write_threads'] = 8 default['opennms']['importer']['requisition_dir'] = "#{default['opennms']['conf']['home']}/etc/imports" default['opennms']['importer']['foreign_source_dir'] = "#{default['opennms']['conf']['home']}/etc/foreign-sources"


Do you use the Remedy integration but aren't the Italian that checked their Remedy config into git? Change all of these attributes then:



Two attributes are available for your availability report running pleasure:

  • storage_location
  • persist_reports

But they aren't very helpful until you add some reports. A LWRP for that is planned.


Change the image format from the default png to gif or jpg (if using jrobin or you like broken images) with node['response_graph']['image_format']. Font sizes can also be changed with node['response_graph']['default_font_size'] and node['response_graph']['title_font_size'] (defaults are 7 and 10 respectively). Setting these attributes to false removes them from the file:

  • icmp
  • avail
  • dhcp
  • dns
  • http
  • http_8080
  • http_8000
  • mail
  • pop3
  • radius
  • smtp
  • ssh
  • jboss
  • snmp
  • ldap
  • strafeping
  • memcached_bytes
  • memcached_bytesrw
  • memcached_uptime
  • memcached_rusage
  • memcached_items
  • memcached_conns
  • memcached_tconns (off by default)
  • memcached_cmds
  • memcached_gets
  • memcached_evictions
  • memcached_threads
  • memcached_struct
  • ciscoping_time

If you changed the count of pings in the strafer polling package to a value higher than 20, you'll also need to define additional colors for the strafeping graph, like default['opennms']['response_graph']['strafeping_colors'][21] = ["#f5f5f5"]. If you want to add a STACK to the graph for another ping number (defaults to 1-4,10,19) add a second color to that attribute's value array, like default['opennms']['response_graph']['strafeping_colors'][21] = ["#f5f5f5","#050505"].


Tobi enthusiasts will want to set some attributes in node['opennms']['rrd'] to switch from jrobin to rrdtool: { "opennms": { "rrd": { "strategy_class": "org.opennms.netmgt.rrd.rrdtool.JniRrdStrategy", "interface_jar": "/usr/share/java/jrrd.jar", "jrrd": "/usr/lib/" } } } TODO: automatically install the appropriate JNI stuff for the target architecture/platform.

You can also change a multitude of queue settings or change the jrobin backend factory, but unless you know what you're doing that's probably a mistake. Look at the template for details if you're curious.

Finally, to turn on the Google protobuf export thing described at, set these attributes accordingly:

default['opennms']['rrd']['usetcp']      = true
default['opennms']['rrd']['tcp']['host'] =
default['opennms']['rrd']['tcp']['port'] = 9100     # Hope that's a JetDirect compatible network interface!


Like other factory configuration files, you can change some settings with attributes:

default['opennms']['rtc']['updaters']                      = 10
default['opennms']['rtc']['senders']                       = 5
default['opennms']['rtc']['rolling_window']                = "24h"
default['opennms']['rtc']['max_events_before_resend']      = 100
default['opennms']['rtc']['low_threshold_interval']        = "20s"
default['opennms']['rtc']['high_threshold_interval']       = "45s"
default['opennms']['rtc']['user_refresh_interval']         = "2m"
default['opennms']['rtc']['errors_before_url_unsubscribe'] = 5


Do you actually populate the building column in assets or site field in provisioning reqs? Change the default site status view name and/or it's definition with these attributes: node['opennms']['site_status_views']['default_view']['name'] and node['opennms']['site_status_views']['default_view']['rows'] where rows is an array of single element hashes (to maintain order) like: [ { "Routers": "Routers" }, { "Switches": "Switches" }, { "Servers": "Servers" } ]


Populate node['opennms']['sms_phonebook']['entries'] with { "hostname": "+PHONE_NUMBER" } keypairs to define an IP address' phone number for the mobile sequence monitor ( A pending LWRP will provide the ability to add the required mobile-sequence elements to make this useful.


Similar to other * files, you can change the image format used in adhoc graphs by setting the attribute node['opennms']['snmp_adhoc_graph']['image_format'] to gif or jpg rather than the default png. Note that the intersection of formats supported by both jrobin and rrdtool is png, though.

etc/ &*

Similar to other * files, you can change the image format used in predefined graphs by setting the attribute node['opennms']['snmp_adhoc_graph']['image_format'] t o gif or jpg rather than the default png. Note that the intersection of formats supported by both jrobin and rrdtool is png, though. You can also set the default and title font sizes like you can in the response graphs. Since these graphs are now split up by manufacturer, you can disable graphs for a manufacturer like you can in snmp-datacollection-config.xml. This example disables Dell graphs:

      "dell_openmanage": false,
      "dell_rac": false

Note that this doesn't delete that file, it merely comments out the reports=... line(s) in the file.

You can also change the default KSC graph by setting node['snmp_graph']['default_ksc_graph'] to the name of a valid graph.


Some attributes are available to configure the 1.9+ SNMP Interface Poller ( Defaults (effectively disabled):

default['opennms']['snmp_iface_poller']['threads']                               = 30
default['opennms']['snmp_iface_poller']['service']                               = "SNMP"
# array of service names that if down cause the SNMP poller to stop polling
default['opennms']['snmp_iface_poller']['node_outage']                           = ["ICMP","SNMP"]
default['opennms']['snmp_iface_poller']['example1']['filter']                    = "IPADDR != ''"
default['opennms']['snmp_iface_poller']['example1']['ipv4_range']['begin']       = ""
default['opennms']['snmp_iface_poller']['example1']['ipv4_range']['end']         = ""
default['opennms']['snmp_iface_poller']['example1']['ipv6_range']['begin']       = "::1"
default['opennms']['snmp_iface_poller']['example1']['ipv6_range']['end']         = "::1"
default['opennms']['snmp_iface_poller']['example1']['interface']['name']         = "Ethernet"
default['opennms']['snmp_iface_poller']['example1']['interface']['criteria']     = "snmpiftype = 6"
default['opennms']['snmp_iface_poller']['example1']['interface']['interval']     = 300000
default['opennms']['snmp_iface_poller']['example1']['interface']['user_defined'] = false
default['opennms']['snmp_iface_poller']['example1']['interface']['status']       = "on"


You can remove either of the default packages or an individual report by setting attributes in node['opennms']['statsd']['PACKAGE_NAME']['REPORT_NAME'] to false. Packages and their reports are:

  • example1
    • top_n
  • response_time_reports
    • top_10_weekly
    • top_10_this_month
    • top_10_last_month
    • top_10_this_year


Whatever this file is for, you can set support.queueId, support.timeout and support.retry with node['opennms']['support']['queueid'], node['opennms']['support']['timeout'] and node['opennms']['support']['retry'].


Similar to site-status-views.xml. Change this to change the default view shown in the Dashboard of the Web UI with the attribute node['opennms']['surveillance_views']['default_view']. Defaults to default. A LWRP is forthcoming that will let you add your own (and set it to default). So this is pretty useless ATM.


Forward alarms to a syslog server. Attributes available in node['opennms']['syslog_north']:

use_defaults       = true
# set above to false if you want the following four attributes to take effect
enabled            = false
nagles_delay       = 1000
batch_size         = 100
queue_size         = 300000
message_format     = "ALARM ID:${alarmId} NODE:${nodeLabel}; ${logMsg}"

Change these attributes in node['opennms']['syslog_north']['destination'] to define the syslog server to send alarm data to:

name            = "localTest"
host            = ""
port            = "514"
ip_protocol     = "UDP"
facility        = "LOCAL0"
max_length      = 1024
send_local_name = true
send_local_time = true
truncate        = false

And set these attributes in node['opennms']['syslog_north']['uei'] to true to limit alarms forwarded to these specific UEIs:



General settings and their defaults:

default['opennms']['syslogd']['port']                   = 10514
default['opennms']['syslogd']['new_suspect']            = false
default['opennms']['syslogd']['parser']                 = "org.opennms.netmgt.syslogd.CustomSyslogParser"
default['opennms']['syslogd']['forwarding_regexp']      = "^.*\s(19|20)\d\d([-/.])(0[1-9]|1[012])\2(0[1-9]|[12][0-9]|3[01])(\s+)(\S+)(\s)(\S.+)"
default['opennms']['syslogd']['matching_group_host']    = 6
default['opennms']['syslogd']['matching_group_message'] = 8
default['opennms']['syslogd']['discard_uei']            = "DISCARD-MATCHING-MESSAGES"

To disable the predefined definitions, set these to false:



Like everything else that has packages, filters, ranges and services, you can override attributes to tune the defaults. See the template and default attributes for details. You can also configure the number of threads with node['opennms']['threshd']['threads'] (default is 5).


Change the RRD repository location or disable threshold groups with the enabled and rrd_repository attributes in node['opennms']['thresholds']['GROUP'] where group can be:

  • mib2
  • cisco
  • hrstorage
  • netsnmp
  • netsnmp_memory_linux
  • netsnmp_memory_nonlinux
  • coffee


Remove one of the default event translations ( by setting an attribute in node['opennms']['translator'] to false. They are:

  • snmp_link_down
  • snmp_link_up
  • hyperic
  • cisco_config_man
  • juniper_cfg_change


Two attributes available: port and new_suspect in node['opennms']['trapd'] that allow you to configure the port to listen for traps on (default 162) and whether or not to create newSuspect events when a trap is received from an unmanaged host (default false).


Change your admin password by setting node['opennms']['users']['admin']['password'] to whatever hashed value of your password OpenNMS uses. Uppercase MD5? In the future we'll generate one during install. You can also change the name and user_comments attributes, I guess.


Disable things that make OpenNMS run better by setting one of these attributes to false:

  • node['opennms']['vacuumd']['statement']:
    • topo_delete_nodes
    • delete_at_interfaces
    • delete_dl_interfaces
    • delete_ipr_interfaces
    • delete_vlans
    • delete_stp_interfaces
    • delete_stp_nodes
    • delete_snmp_interfaces
    • delete_nodes
    • delete_ip_interfaces
    • delete_if_services
    • delete_events
  • node['opennms']['vacuumd']['automations']:
    • cosmic_clear
    • clean_up
    • full_clean_up
    • gc
    • full_gc
    • unclear
    • escalation
    • purge_statistics_reports
    • create_tickets
    • create_critical_ticket
    • update_tickets
    • close_cleared_alarm_tickets
    • clear_alarms_for_closed_tickets
    • clean_up_rp_status_changes
    • maintenance_check (default false)
    • add_missing_access_points_to_table (default false)
    • update_access_points_table (default false)
    • clean_up_access_points_table (default false)
  • node['opennms']['vacuumd']['triggers']:
    • select_closed_ticket_state_for_problem_alarms
    • select_null_ticket_state_alarms
    • select_critial_open_alarms
    • select_not_null_ticket_state_alarms
    • select_cleared_alarm_with_open_ticket_state
    • select_suspect_alarms
    • select_cleared_alarms
    • select_resolvers
    • select_expiration_maintenance (default false)
    • select_access_points_missing_from_table (default false)
  • node['opennms']['vacuumd']['actions']:
    • acknowledge_alarm
    • update_automation_time
    • escalate_alarm
    • reset_severity
    • garbage_collect_7_4
    • garbage_collect_8_1 (default false)
    • full_garbage_collect_7_4
    • full_garbage_collect_8_1 (default false)
    • delete_past_cleared_alarms
    • delete_all_past_cleared_alarms
    • clear_problems
    • clear_closed_ticket_alarms
    • delete_purgeable_statistics_reports
    • do_nothing_action
    • clean_up_rp_status_changes
    • maintenance_expiration_warning (default false)
    • add_access_point_to_table (default false)
    • update_access_points_table (default false)
    • clean_up_access_points_table (default false)
  • node['opennms']['vacuumd']['auto_events']:
    • escalation_event
  • node['opennms']['vacuumd']['action_events']:
    • create_ticket
    • update_ticket
    • close_ticket
    • event_escalated
    • maintenance_expiration_warning (default false)

You can also change the frequency in which vacuumd runs by setting node['opennms']['vacuumd']['period']. Default is 86400000 (1 day).


Another web UI XML file, this one controls which categories are displayed in the availability box on the main landing page. Once a LWRP exists you'll be able to add sections, but until then you can disable any of the existing categories by setting one of these attributes in node['opennms']['web_console_view'] to false:

  • network_interfaces
  • web_servers
  • email_servers
  • dns_dhcp_servers
  • db_servers
  • jmx_servers
  • other_servers


Similar to other *datacollection-config.xml files, you can change the RRD repository, step, RRA definitions and disable default vmware-cim-groups.


Similar to other *datacollection-config.xml files, you can change the RRD repository, step, RRA definitions and disable default vmware-groups.


Similar to other *datacollection-config.xml files, you can change the RRD repository, step, RRA definitions and disable default wpms.


Note that XML datacollection is a separate plugin package in it's own recipe. Similar to other *datacollection-config.xml files, you can change the RRD repository, step, RRA definitions and disable default xml-collections.


Factory settings can be changed with these attributes in node['opennms']['xmlrpcd']:

['max_event_queue_size']            = 5000
['external_servers']['retries']     = 3
['external_servers']['elapse_time'] = 15000
['external_servers']['url']         = "http://localhost:8000"
['external_servers']['timeout']     = 0

Eventually a LWRP will exist to add more than 1 external_servers.

To change the list of events sent via XML-RPC, override the node['opennms']['xmlrpcd']['base_events'] attribute with an array of UEI strings.


Configure notifications to be sent via XMPP (aka Jabber, GTalk) with these attributes in node['opennms']['xmpp']:

  • server
  • service_name
  • port
  • tls
  • sasl
  • self_signed_certs
  • truststore_password
  • debug
  • user
  • pass


Apache 2.0


David Schlenk (


Please feel free to fork and send me pull requests! The focus of my work will initially be on templates for configuration files that modify the default configuration and LWRPs to add new elements to configuration files.

There's a test kitchen suite for every main and example recipe. The default test kitchen config uses the vagrant driver, but I actually use the openstack driver (overriding the default with .kitchen.local.yml), so if the default config doesn't work that's why and I'm so, so sorry.

Dependent cookbooks

build-essential >= 0.0.0
java >= 0.0.0
hostsfile >= 0.0.0
yum >= 0.0.0

Contingent cookbooks

There are no cookbooks that are contingent upon this one.

No quality metric results found