Connector Configuration

Providing Configuration

For information about about how the connector detects configuration properties, see Spring Boot: Externalized Configuration.

Converting Canonical Spring Property Names to Environment Variables

For information about converting the Spring property names to environment variables, see the Spring documentation.

Spring Profiles

If multiple configuration files exist within the same configuration directory for use in different environments (development, production, etc.), use Spring profiles.

Using Spring profiles allow you to define different application property files under the same directory using the filename format, application-{profile}.yml.

For example:

application.yml: The properties in non-specific files that always apply. Its properties are overridden by the properties defined in profile-specific files.
application-dev.yml: Defines properties specific to the development environment.
application-prod.yml: Defines properties specific to the production environment.

Individual profiles can then be enabled by setting the spring.profiles.active property. See Spring Boot: Profile-Specific Files for more information and an example.

Configuring Locations to Find Spring Property Files

By default, the connector detects any Spring property files as described in the Spring Boot’s default locations.

If you want to add additional locations, add --spring.config.additional-location=file:<custom-config-dir>.
If you want to exclusively use the locations that you’ve defined and ignore Spring Boot’s default locations, add --spring.config.location=optional:classpath:/,optional:classpath:/config/,file:<custom-config-dir>.

For example of the command to add additional locations, see Quick Start: Running the Connector Using the Command Line.

For more information about configuring locations to find Sprint property files, see Spring Boot documentation.

If you want configuration files for multiple, different connectors within the same config directory for use in different environments (such as, development, production, etc.), we recommend that you use Spring Boot Profiles instead of child directories. For example:

Set up your configuration like this:
- config/application-prod.yml
- config/application-dev.yml
Do not do this:
- config/prod/application.yml
- config/dev/application.yml

Child directories are intended to be used for merging configuration from multiple sources of confiugration properties. For more information and an example of when you might want to use multiple child directories to compose your application’s configuration, see the Spring Boot documentation.

Obtaining Build Information

For connector versions 2.0.0 and later, build information that includes the version, build date, time, and description is enabled by default using the Spring Boot Actuator Info Endpoint. By default, a connector only shares information related to its build.

The following example shows the structure of the output data:

{
"build": {
"version": "<connector version>",
"artifact": "<connector artifact>",
"name": "<connector name>",
"time": "<connector build time>",
"group": "<connector group>",
"description": "<connector description>",
"support": "<support information>"
}
}

If you want to exclude build data from the output of the info endpoint, set management.info.build.enabled to false.

Alternatively, if you want to disable the info endpoint entirely, you can remove info from the list of endpoints specified in management.endpoints.web.exposure.include.

Spring Configuration Options

This connector packages use Spring configuration options to customize functionality that includes Spring Cloud Binder for Solace PubSub+ and other common Spring Boot libraries.

Using the Spring Cloud Binder for Solace PubSub+

An important piece for customizing the Spring configuration options for Self-Managed Connectors is the Spring Cloud Binder for Solace PubSub+. There are configuration options available that you may find useful to configure the connector such as:

endpointType, which is a consumer configuration option to configure the type of endpoint (queue or topic endpoint) that consumers read messages from
selector, which is a consumer configuration option to configure an endpoint's message selector
batchMaxSize, which is the maximum number of messages per batch. . If a consumer polls for messages and none are available, a partial batch is created. It's important to note that all messages in the batch are rejected if both the following conditions exist:
- The batchMaxSize is more than 1.
- An error occurs during the processing of any message within the batch.
For example, if you have a batch of 50 messages and one message fails, all 50 messages in the batch are rejected.
producer configuration options to publish messages within transacted sessions
authentication schemes to connect to a PubSub+ event broker, such as basic authentication (default) , client certificate, or OAuth 2.0

For more information about:

Other configuration options available for the Spring Cloud Binder for Solace PubSub+ and detailed information for the configuration options in the list above, see the Solace Binder Configuration Options.
The latest, supported features for the Spring Cloud Binder for Solace PubSub+, see the Spring Cloud Binder for Solace PubSub+ release notes.

Using Other Spring Boot Libraries

The following are other common references to Spring Boot libraries that you can use to configure your connector that you may find useful:

JMS Binder Configuration Options

These properties are to be prefixed with jms-binder.


Config Option	Type	Valid Values	Default Value	Description
`health-check.interval`	`long`	> 0	10000	Interval (in ms) between reconnection attempts while health status is `RECONNECTING`
`healthcheck.reconnectattempts-until-down`	`long`	>=0	10	The number of reconnection attempts until the JMS binder transitions from `RECONNECTING` to `DOWN`. A value of 0 means unlimited number of attempts which means that the binder would never transition to the `DOWN` state.
`jndi.context`	`java.util. Properties`	`key/value pairs`		Standard JNDI properties. See JNDI Context section for details.
`jndi.connectionfactory.name`	`String`			The connection factory’s JNDI name used for the lookup
`jndi.connectionfactory.user`	`String`			The user to authenticate with the JMS compliant broker.
`jndi.connectionfactory.password`	`String`			The password to authenticate with the JMS compliant broker.

The properties in the table above are complementary to the properties described in Configuring Connection Details.

JMS Consumer Options

The following configuration options are available to JMS consumers. Options within the same table share the same prefix.

The following options in Config Option are prefixed with spring.cloud.stream.jms.bindings.<bindingName>.consumer.


Config Option	Type	Valid Values	Default Value	Description
`batch-max-size`	`int`	`>= 1`	`255`	The maximum number of messages that can be grouped together in a single batch. If a consumer polls for messages and none are available, a partial batch is created. You can configure the batch size to one to disable batching. It's important to note that all messages in the batch are rejected if both the following conditions exist: The `batchMaxSize` is more than 1. An error occurs during the processing of any message within the batch. For example, if you have a batch of 50 messages and one message fails, all 50 messages in the batch are rejected.
`transacted`	`boolean`	`true` or `false`	`true`	Specifies whether messages are received within a local transaction. When set to `true`, it indicates that the JMS consumer reads messages within a local transaction and commits the transaction when the batch has been successfully processed. Set to `false` to disable transactions. We recommend that you disable transactions because single message transactions (i.e., when `batch-max-size=1`) lower your performance significantly.
`destination-type`	`String`	`(queue\|topic\|unknown)`	`unknown`	The type of destination where messages are consumed from. queue The destination value is assumed to be a physical queue and no JNDI lookup is done. topic The destination value is assumed to be a physical topic and no JNDI lookup is done. This option requires `durable-subscription-name` to also be set. unknown The destination value is assumed to be a JNDI name. The actual destination name is only known after a successful lookup. A valid JNDI context must be configured via `jms-binder.jndi.context`.
`durablesubscription-name`	`String`	A valid subscription.	None.	The name of the shared durable subscription to consume from. The subscription is created on the broker if it doesn’t already exist. Applies, and is mandatory, when `destination-type` is or resolves to a topic.

The following options in Config Option are prefixed with spring.cloud.stream.bindings.<bindingName>.consumer.


Config Options	Type	Valid Values	Default Value	Description
`concurrency`	`int`	> 0	1	The number of concurrent consumers to create.

If a configuration option applies to all JMS input bindings, it can be prefixed with either:

spring.cloud.stream.jms.default.consumer if the option is from the first table.
spring.cloud.stream.default.consumer if the option is from the second table.

The mentioned classes are convenient way to assign a configuration to all JMS input bindings.

JMS Producer Options

The following configuration options are available to JMS producers.

The following options in Config Option are prefixed with spring.cloud.stream.jms.bindings.<bindingName>.producer.


Config Option	Type	Valid Values	Default Value	Description
`destination-type`	`String`	`(queue\|topic\|unknown)`	`unknown`	The type of destination where messages are published to. queue The destination value is assumed to be a physical queue and no JNDI lookup is done. topic The destination value is assumed to be a physical topic and no JNDI lookup is done. unknown The destination value is assumed to be a JNDI name. The actual destination name is only known after a successful lookup. A valid JNDI context must be configured via `jms-binder.jndi.context`.
`transacted`	`boolean`	`true` or `false`	`true`	Specifies whether the JMS producer publishes messages from a received batch using a local transaction. Setting transacted to `true` provides duplicate protection in case of producer failures. Set to `false` to disable transactions.

If a configuration option applies to all JMS output bindings, it can be prefixed with spring.cloud.stream.jms.default.producer. This prefix is a convenient way to assign a configuration to all JMS output bindings.

Connector Configuration Options

The following table lists the configuration options. The following options in Config Option are prefixed with solace.connector.:

Config Option	Type	Valid Values	Default Value	Description
`management.leader- election.fail-over.max-attempts`	`int`	`> 0`	`3`	The maximum number of attempts to perform a fail-over.
`management.leader- election.fail-over.back-off-initial-interval`	`long`	`> 0`	`1000`	The initial interval (milliseconds) to back-off when retrying a fail-over.
`management.leader- election.fail-over.back-off-max-interval`	`long`	`> 0`	`10000`	The maximum interval (milliseconds) to back-off when retrying a fail-over.
`management.leader-election.fail-over.back- off-multiplier`	`double`	`>= 1.0`	`2.0`	The multiplier to apply to the back-off interval between each retry of a fail-over.
`management.leader- election.mode`	`enum`	`(standalone\|active_active\| active_standby)`	`standalone`	The connector’s leader election mode. standalone: A single instance of a connector without any leader election capabilities. active_active: A participant in a cluster of connector instances where all instances are active. active_standby: A participant in a cluster of connector instances where only one instance is active (i.e. the leader), and the others are standby.
`management.queue`	`string`	`any`	`null`	The management queue name.
`management.session.*`		See Spring Boot Auto-Configuration for the Solace Java API		Defines the management session. This has the same interface as that used by `solace.java.*`. See Spring Boot Auto-Configuration for the Solace Java API for more information.
`security.enabled`	`boolean`	`(true\|false)`	`true`	If `true`, security is enabled. Otherwise, anyone has access to the connector’s endpoints.
`security.csrf-enabled` (Only for versions earlier than 2.0.0)	`boolean`	`(true\|false)`	`true`	If true, CSRF protection is enabled. Makes sense only if `solace.connector.security.enabled` is true.
`security.users[<index>].name`	`string`	any	null	The name of the user.
`security.users[<index>].password`	`string`	any	null	The password for the user.
`security.users[<index>].roles`	`list<string>`	admin	empty list (i.e. read- only)	The list of roles that the specified user has. It has read-only access if no roles are returned.

Workflow Configuration Options

These configuration options are defined under the following prefixes:

solace.connector.workflows.<workflow-id>.: If the options support per-workflow configuration and the default prefixes.
solace.connector.default.workflow.: If the options support default workflow configuration.

Config Option	Applicable Scopes	Type	Valid Values	Default Value	Description
`enabled`	Per-Workflow	`boolean`	`(true\|false)`	`false`	If `true`, the workflow is enabled.
`transform-headers.expressions`	Per-Workflow Default	`Map<string,string>`	Key: A header name. Value: A SpEL string that accepts `headers` as parameters.	`empty map`	A mapping of header names to header value SpEL expressions. The SpEL context contains the headers parameter that can be used to read the input message’s headers.
`acknowledgment.publish-async`	Per- Workflow Default	`boolean`	`(true\|false)`	`false`	If `true`, publisher acknowledgment processing is done asynchronously. The workflow’s consumer and producer bindings must support this mode, otherwise the publisher acknowledgments are processed synchronously regardless of this setting.
`acknowledgment.back- pressure-threshold`	Per-Workflow Default	`int`	`>= 1`	`255`	The maximum number of outstanding messages with unresolved acknowledgments. Message consumption is paused when the threshold is reached to allow for producer acknowledgments to catch up.
`acknowledgment.publish-timeout`	Per-Workflow Default	`int`	`>= -1`	`600000`	The maximum amount of time (in milliseconds) to wait for asynchronous publisher acknowledgments before considering a message as failed. A value of `-1` means to wait indefinitely for publisher acknowledgments.

Provide feedback