Cafienne Engine Configuration
Introduction
This page gives an overview of the various configuration options of the Cafienne Engine.
Cafienne Engine runs on the Akka actor system and is typically hosted in a configuration of Docker containers.
docker-compose.yml, local.conf & environment variables
Both Docker and Akka have a wide range of configuration possibilities.
Both languages also support the use of environment variables.
An example configuration that combines all these options can be found in the Cafienne Demo environment of the getting-started repository in GitHub.
Various persistence configurations
The Cafienne Engine requires two types of storage, based on the CQRS principle. The configuration for the event journal is done through that standard akka persistence configuration.
| storage | |
|---|---|
| Event Journal | Events that are generated during the lifecycle of a case are stored in the Event Journal. |
| Query DB | In order to have fast retrieval of case information, projections of events are stored in the Query database. |
Note: include the following configuration when using Microsoft SQL Server as event journal
This setting is required to avoid full table scans while inserting events.
url = "jdbc:sqlserver://...;DatabaseName=event-journal;sendStringParametersAsUnicode=false;
- Example configuration for Cassandra
The Cassandra configuration uses Cassandra as Event Journal, and PostgreSQL as Query DB. - Example configuration for PostgreSQL
- Example configuration for Microsoft SQL Server
Cafienne specific settings
Cafienne Engine specific settings can be added to the Akka configuration file. These settings include the following:
cafienne {
# Engine wide platform settings
platform {
# Platform has owners that are allowed to create/disable/enable tenants
# This property specifies the set of user-id's that are owners
# This array may not be empty.
owners = ["admin"]
owners = ${?CAFIENNE_PLATFORM_OWNERS}
# Default tenant will be used when a user does not provide the tenant as a parameter to
# the API call (e.g. in StartCase). When the user is member of only one tenant,
# then that tenant will be submitted along with the StartCase command.
# If the user belongs to multiple tenants, then this default-tenant option will be passed.
default-tenant = "world"
default-tenant = ${?CAFIENNE_PLATFORM_DEFAULT_TENANT}
# bootstrap-tenants holds a reference to zero or more json (or yaml) files that have default tenant information.
# Each file is should contain information for one tenant (tenant name, users and owners).
# During launch of the case engine, the files will be scanned and a CreateTenant command is sent
# into the system, thereby setting up one or more default tenants.
# If the bootstrap-tenants property is not filled and the default-tenant has a value, the system will search for
# a file that holds the default tenant name plus either a .conf, .json, .yml or .yaml extension.
# E.g. in case default-tenant = 'world', the system would search for existence in the following order:
# - 'world.conf'
# - 'world.json'
# - 'world.yml'
# - 'world.yaml'
# If none of these files are found, the bootstrap attempt will be skipped.
bootstrap-tenants = ["world.conf"]
}
engine {
# Properties for sending tasks of type Mail
mail-service {
# Here you can fill any regular jakarta.mail properties
# All properties mentioned here are passed into the connection with the mail server
mail.host = mailcatcher
mail.smtp.port = 1025
}
# more settings can be found in the getting-started configuration
}
api {
bindhost = "localhost"
bindport = 2027
security {
###############################################################################
## Below settings can be used to configure multiple OIDC services ##
## that ships with the docker images in the cafienne repository named ##
## getting-started. ##
###############################################################################
# configuration settings for OpenID Connect
oidc = [{
### This one has dynamic resolution (only a issuer needs to be set)
issuer = "http://localhost:5556/dex"
}, {
### This one seems to have dynamic resolution ...
### but since the key-url is filled, that will take precedence (at this moment)
connect-url = "http://localhost:2377"
token-url = "http://localhost:2377/token"
key-url = "http://localhost:2377/keys"
authorization-url = "http://localhost:2377/auth"
issuer = "http://localhost:2377"
}, {
### This one will be skipped, unless the environment variables are filled.
connect-url = ${?CAFIENNE_OIDC_CONNECT_URL}
token-url = ${?CAFIENNE_OIDC_TOKEN_URL}
key-url = ${?CAFIENNE_OIDC_KEY_URL}
authorization-url = ${?CAFIENNE_OIDC_AUTHORIZATION_URL}
issuer = ${?CAFIENNE_OIDC_ISSUER}
}]
}
definitions {
# The case engine supports various ways to list, load and deploy case definitions.
}
actor {
# the seconds of wait time before a response to a command is expected
# by the akka http command routes
ask-timeout = 60
# the seconds of idle time after which a case actor is removed from akka memory
# if the case has not received new commands after the specified number of seconds,
# the case engine will ask akka to remove the case from memory to avoid memory leaks.
idle-period = 600
}
# This setting tells cafienne which journal to use for reading events.
# If omitted, cafienne will try to guess the read journal, based on the akka settings
read-journal = "jdbc-read-journal"
query-db {
profile = ${?QUERY_DB_PROFILE}
db {
driver = ${?QUERY_DB_DRIVER}
user = ${?QUERY_DB_USER}
password = ${?QUERY_DB_PASSWORD}
numThreads = 10
connectionTimeout = 5000
validationTimeout = 5000
url = ${?QUERY_DB_URL}
}
# Configuration options handling exceptions that may occur while reading
# the event streams that populate the query-db tables
# See also https://doc.akka.io/docs/akka/current/stream/stream-error.html#restart-with-backoff
restart-stream {
min-back-off = 500ms
max-back-off = 30s
random-factor = 0.20
max-restarts = 20
max-restarts-within = 5m
}
}
}