Skip to main content



The config.rb file is a replacement for the knife.rb file. The config.rb file has identical settings and behavior to the knife.rb file. Chef Infra Client looks first for the presence of the config.rb file and if it is not found, then looks for the knife.rb file.

A config.rb file is used to specify configuration details for knife.

A config.rb file:

  • Is loaded every time the knife executable is run
  • Is not created by default
  • Is located by default at ~/.chef/config.rb (macOS and Linux platforms) or c:\Users\username\.chef\config.rb (Windows platform), use the --config option from the command line to change this location
  • Will override the default configuration when a config.rb file exists at the default path or the path specified by the --config option


When running Windows, the config.rb file is located at %HOMEDRIVE%:%HOMEPATH%\.chef (e.g. c:\Users\<username>\.chef).


This configuration file has the following settings:

The path to a template file to be used during a bootstrap operation.
The URL of the Chef Infra Server. For example:
chef_server_url 'https://localhost/organizations/ORG_NAME'


If changes need to be made to any global end points like user or user keys, use knife exec with the --server-url flag to set chef_server_url to https://localhost/.
Enable chef-zero. This setting requires local_mode to be set to true. Default value: false. For example:
chef_zero.enabled true
The port on which chef-zero is to listen. Default value: 8889. For example:
chef_zero[:port] 8889
A directory that contains additional configuration scripts to load for Chef Infra Client.
The location of the file that contains the client key, as an absolute path. Default value: /etc/chef/client.pem. For example:
client_key '/etc/chef/client.pem'
The name of the copyright holder. This option places a copyright notice that contains the name of the copyright holder in each of the pre-created files. If this option is not specified, a copyright name of “COMPANY_NAME” is used instead; it can be modified later.
The email address for the individual who maintains the cookbook. This option places an email address in each of the pre-created files. If not specified, an email name of “YOUR_EMAIL” is used instead; this can be modified later.
The type of license under which a cookbook is distributed: apachev2, gplv2, gplv3, mit, or none (default). This option places the appropriate license notice in the pre-created files: Apache v2.0 (for apachev2), GPL v2 (for gplv2), GPL v3 (for gplv3), MIT (for mit), or license 'Proprietary - All Rights Reserved (for none). Be aware of the licenses for files inside of a cookbook and be sure to follow any restrictions they describe.
The Chef Infra Client sub-directory for storing cookbooks. This value can be a string or an array of file system locations, processed in the specified order. The last cookbook is considered to override local modifications. For example:
cookbook_path [
The minimum required version of data bag encryption. Possible values: 1 or 2. When the machines in an organization are running Chef Infra Client 11.6 (or higher), it is recommended that this value be set to 2. For example:
data_bag_encrypt_version 2
Allows OpenSSL to enforce FIPS-validated security during a Chef Infra Client run. Set to true to enable FIPS-validated security.

The following operating systems are supported:

  • Red Hat Enterprise Linux
  • Oracle Enterprise Linux
  • CentOS
  • Ubuntu (with Ubuntu Advantage subscription)
  • Windows
Run Chef Infra Client in local mode. This allows all commands that work against the Chef Infra Server to also work against the local chef-repo. For example:
local_mode true

The name of the node. This may be a username with permission to authenticate to the Chef Infra Server or it may be the name of the machine from which knife is run. For example:

node_name 'user_name'


node_name 'machine_name'

A comma-separated list of URLs that do not need a proxy. Default value: nil. For example:

no_proxy 'localhost,, *, *'
New in 14.2 Use ssh-agent to authenticate. When using this option, specify the location of the public key in client_key. Default value: false. Ensure the public key is in PKCS#1 format. You can convert an OpenSSH public key using ssh-keygen. For example:
ssh-keygen -f -e -m pem > key.pem
The amount of time (in seconds) to wait for an SSH connection time out.
Set the verify mode for HTTPS requests.
  • Use :verify_none to do no validation of SSL certificates.
  • Use :verify_peer to do validation of all SSL certificates, including the Chef Infra Server connections, S3 connections, and any HTTPS remote_file resource URLs used in a Chef Infra Client run. This is the recommended setting.

Depending on how OpenSSL is configured, the ssl_ca_path may need to be specified. Default value: :verify_peer.


Split the Tmux window. Default value: false.


The name of the chef-validator key that is used by Chef Infra Client to access the Chef Infra Server during the initial Chef Infra Client run when not using validatorless bootstrapping. For example:

validation_client_name 'chef-validator'

The location of the file that contains the key used when a Chef Infra Client is registered with a Chef Infra Server. A validation key is signed using the validation_client_name for authentication. Default value: /etc/chef/validation.pem. For example:

validation_key '/etc/chef/validation.pem'
Verify the SSL certificate on the Chef Infra Server. When true, Chef Infra Client always verifies the SSL certificate. When false. Chef Infra Client uses the value of ssl_verify_mode to determine if the SSL certificate requires verification. Default value: false.
Append cookbook versions to cookbooks. Set to false to hide cookbook versions: cookbooks/apache. Set to true to show cookbook versions: cookbooks/apache-1.0.0 and/or cookbooks/apache-1.0.1. When this setting is true, knife download downloads ALL cookbook versions, which can be useful if a full-fidelity backup of data on the Chef Infra Server is required. For example:
versioned_cookbooks true
Sets the default value of log_level in the client.rb file of the node being bootstrapped. Possible values are :debug, :info, :warn, :error and :fatal. For example:
config_log_level :debug
Sets the default value of log_location in the client.rb file of the node being bootstrapped. Possible values are /path/to/log_location, STDOUT, STDERR, :win_evt and :syslog. For example:
config_log_location "/path/to/log_location"   # Please make sure that the path exists

Proxy Settings

In certain situations the proxy used by the Chef Infra Server requires authentication. In this situation, three settings must be added to the configuration file. Which settings to add depends on the protocol used to access the Chef Infra Server: HTTP or HTTPS.

If the Chef Infra Server is configured to use HTTP, add the following settings:

The proxy server for HTTP connections. Default value: nil. For example:
http_proxy ''
The user name for the proxy server when the proxy server is using an HTTP connection. Default value: nil.
The password for the proxy server when the proxy server is using an HTTP connection. Default value: nil.

HTTPS Proxy Settings (such as the hosted Chef Infra Server)

The proxy server for HTTPS connections. (The hosted Chef Infra Server uses an HTTPS connection.) Default value: nil.
The user name for the proxy server when the proxy server is using an HTTPS connection. Default value: nil.
The password for the proxy server when the proxy server is using an HTTPS connection. Default value: nil.

Settings for No-proxy URLs

A comma-separated list of URLs that do not need a proxy. Default value: nil.

.d Directories

Chef Infra Client supports reading multiple configuration files by putting them inside a .d configuration directory. For example: /etc/chef/client.d. All files that end in .rb in the .d directory are loaded; other non-.rb files are ignored.

.d directories may exist in any location where the client.rb, config.rb, or solo.rb files are present, such as:

  • /etc/chef/client.d
  • /etc/chef/config.d
  • ~/chef/solo.d

(There is no support for a knife.d directory; use config.d instead.)

For example, when using knife, the following configuration files would be loaded:

  • ~/.chef/config.rb
  • ~/.chef/config.d/company_settings.rb
  • ~/.chef/config.d/ec2_configuration.rb
  • ~/.chef/config.d/old_settings.rb.bak

The old_settings.rb.bak file is ignored because it is not a configuration file. The config.rb, company_settings.rb, and ec2_configuration files are merged together as if they are a single configuration file.


If multiple configuration files exists in a .d directory, ensure that the same setting has the same value in all files.

Optional Settings

In addition to the default settings in a config.rb file, there are other subcommand-specific settings that can be added:

  1. A value passed via the command-line
  2. A value contained in the config.rb file
  3. The default value

A value passed via the command line overrides a value in the config.rb file; a value in a config.rb file overrides a default value. Before adding any settings to the config.rb file:

  • Verify the settings by reviewing the documentation for the knife subcommands and/or knife plugins
  • Verify the use case(s) your organization has for adding them

Also note that:

  • Custom plugins can be configured to use the same settings as the core knife subcommands
  • Many of these settings are used by more than one subcommand and/or plugin
  • Some settings are included only because knife checks for a value in the config.rb file

To add settings to the config.rb file, use the following syntax:

knife[:setting_name] = value

where value may require quotation marks (’ ‘) if that value is a string. For example:

knife[:ssh_port] = 22
knife[:bootstrap_template] = 'ubuntu14.04-gems'
knife[:bootstrap_version] = ''
knife[:bootstrap_proxy] = ''

Frequently Used

Some optional config.rb settings are used often, such as the template file used in a bootstrap operation. The frequency of use of any option varies from organization to organization, so even though the following settings are often added to a config.rb file, they may not be the right settings to add for every organization:

The proxy server for the node that is the target of a bootstrap operation.
The path to a template file to be used during a bootstrap operation.
The version of Chef Infra Client to install.
The $EDITOR that is used for all interactive commands.
The SSH tunnel or gateway that is used to run a bootstrap action on a machine that is not accessible from the workstation. Adding this setting can be helpful when a user cannot SSH directly into a host.
The SSH port.

Additional SSH Settings

Other SSH-related settings that are sometimes helpful when added to the config.rb file:

Enable SSH agent forwarding.
The attribute used when opening an SSH connection.
The SSH password. This can be used to pass the password directly on the command line. If this option is not specified (and a password is required) knife prompts for the password.
The SSH user name.

Data Bag Settings

Some organizations choose to have all data bags use the same secret and secret file, rather than have a unique secret and secret file for each data bag. To use the same secret and secret file for all data bags, add the following to config.rb:

The encryption key that is used for values contained within a data bag item.
The path to the file that contains the encryption key.


Review the full list of optional settings that can be added to the config.rb file. Many of these optional settings should not be added to the config.rb file. The reasons for not adding them can vary. For example, using --yes as a default in the config.rb file causes knife to always assume that “Y” is the response to any prompt, which may lead to undesirable outcomes. Other settings, such as --hide-healthy(used only with the knife status subcommand) or --bare-directories (used only with the knife list subcommand) probably aren’t used often enough (and in the same exact way) to justify adding them to the config.rb file. In general, if the optional settings are not listed on the main config.rb page, then add settings only after careful consideration. Do not use optional settings in a production environment until after the setting’s performance has been validated in a safe testing environment.
Edit this page on GitHub

Thank you for your feedback!


Search Results