AI Asset Manager configurations

AI Asset Manager has many configurable settings, all of which come with working and pre-configured values when you install the application. However, if necessary, they can be reconfigured through the IEM configuration mechanism. The deployable app configuration file should list only the properties whose default value is to be changed.

This section describes the default values for the options used for connecting to Cloud providers and Edge apps.

The default values listed below are extended by explanatory comment lines that begin with the "#" character.

These lines can optionally be included in the configuration file but do not affect the application.

## Request timeout settings

##! Time interval for the system to serve a request.

##! (Hence the database connection is established in the context of the request,

##! once the timeout expires the database context is terminated.)

##! Default: 10 minutes

request.timeout=10m

## IEM response timeout setting

##! Time interval to wait for IEM response.

##! Default: 10 seconds

iem.request.timeout=10s

## log.x.level can specify the right amount of log details for the given domain

## Supported log levels: "TRACE / DEBUG / INFO / WARN / ERROR / PANIC / FATAL"

##! The default is ERROR. That means only error messages are displayed in the log.

log.general.level=ERROR

log.database.level=ERROR

log.mqtt.level=ERROR

## Rules for identifying compatible deployment target applications

##! - "apps" is the list of applications that must be installed on the device

##! - "targetAppIndex" identifies the embedder among these applications which is the target of the deployments (defaults to 0 if ommited)

##! - "targetAppType" determines how AI Asset Manager communicates with the target. Target applications having similar

##! REST APIs form a target type, like "AI Inference Server" and all its variants (3 pipelines, GPU, etc.)

##! Supported Target Application Types (targetAppType) with their code names:

##! - "aiis": "AI Inference Server" and its variants

##! Default rules are AI Inference Server and its variants that means AI Asset Manager considers only those devices

##! compatible that has the given application variant installed.

##! In order to set compatibility with another application, additional rules can be defined.

##! E.g.: deviceRules.ruleName={"apps": , "targetAppType": "codeOfTheTargetApplicationType"}

deviceRules.aiInferenceServer={"targetAppIndex": 0, "apps": , "targetAppType": "aiis"}

deviceRules.aiInferenceServer3Pipelines={"targetAppIndex": 0, "apps": , "targetAppType": "aiis"}

deviceRules.aiInferenceServerGPU={"targetAppIndex": 0, "apps": , "targetAppType": "aiis"}

Starting with AI Asset Manager v2.0.0 with the introduction of the Connection Upgrade feature, the above-mentioned "targetAppType" attribute must be specified in the device rule for all features of the New Communication method to be enabled.

AI Asset Manager v2.0.0 and above, comes out of the box with the above-mentioned AI Inference Server types and the needed "targetAppType" attributes. If a custom application is defined with a device rule or the default configuration is overridden, then this attribute must be specified in the device rule.