Configuration
See full configuration here
Configuration files are in HOCON format
Network configuration
Property name | Default value | Meaning |
---|---|---|
mist.cluster.host |
0.0.0.0 | default communication host |
mist.cluster.port |
2551 | default communication port |
mist.http.host |
0.0.0.0 | default http host |
mist.http.port |
2004 | default http port |
mist.log-service.host |
0.0.0.0 | default http host |
mist.log-service.port |
2005 | default http port |
Also it’s possible to use auto
value for host configuration keys - mist automactically infers the best matched host address.
mist.cluster.host = "auto"
It’s useful in case when mist works in environment where we need to set up public visible address. For example: running mist-worker in separate docker container, using kubernetes cluster for spark
Database configuration
To store information about jobs and their statuses Mist uses H2 as database. From v1.1.2 it’s also possible to use PostgreSQL. Example:
mist.db {
poolSize = 32
driverClass = "org.postgresql.Driver"
jdbcUrl = "jdbc:postgresql:mist"
username = "..."
password = "..."
}
By default, automatic db migration is enabled. It means that the provided user should be allowed to create tables. In case if it’s not possible you can disable it and maintain your database manually using these migration scripts.
mist.db {
migration = off
}
Worker configuration
Property name | Default value | Meaning |
---|---|---|
mist.workers.runner-init-timeout |
2 minutes | Worker creation timeout |
mist.workers.ready-timeout |
2 minutes | Spark context creation timeout |
mist.workers.runner |
local | How mist spawn workers |
Runners
There are several ways how mist spawn workers:
local
- just localsubmit-submit
on the same nodedocker
- Spawn worker in separated docker containermanual
- custom script for worker spawning - example.
See additional configuration for docker
and manual
runners bellow:
Docker
Property name | Default value | Meaning |
---|---|---|
workers.docker.host |
unix:///var/run/docker.sock | Docker server adress |
workers.docker.image |
hydrosphere/mist:1.1.3-2.3.0” | Mist docker image |
workers.docker.mist-home |
/usr/share/mist | Path to mist home inside image |
workers.docker.spark-home |
/usr/share/spark | Path to spark home inside image |
workers.docker.network-type |
net | Docker network driver. Allowed values: net , bridge , auto-master - clone master network settings if it’s runned inside container |
worker.docker.auto-master-network.container.id |
Mist master docker container id (usage) |
Manual
Property name | Default value | Meaning |
---|---|---|
workers.manual.startCmd |
path to start script | |
workers.manual.stopCmd |
path to stop script |
Enable async interfaces to receive events
Kafka:
Property name | Default value | Meaning |
---|---|---|
mist.kafka.on |
false | Turning on/off support of async jobs with kafka |
mist.kafka.host |
localhost | Broker host |
mist.kafka.port |
9092 | Broker port |
mist.kafka.subscribe-topic |
no default value | Topic Mist listens to for incoming requests |
mist.kafka.publish-topic |
no default value | Topic Mist writes response into |
MQTT:
Property name | Default value | Meaning |
---|---|---|
mist.mqtt.on |
false | Turning on/off support of async jobs with MQTT |
mist.mqtt.host |
localhost | MQTT server host |
mist.mqtt.port |
1883 | MQTT server port |
mist.mqtt.subscribe-topic |
no default value | Topic Mist listens to for incoming requests |
mist.mqtt.publish-topic |
no default value | Topic Mist writes response into |
Enable Kerberos
Example:
mist {
security {
enabled = true
keytab = ${path_to_keytab}
principal = ${your_principal}
interval = 30 minutes # default 1 hour
}
}
Default context
To override settings for default
context use prefix mist.context-defaults
- (master.conf)[https://github.com/Hydrospheredata/mist/blob/master/mist/master/src/main/resources/master.conf#L73].