Segregate Scheduler

Overview

tsuru uses schedulers to chooses which node an unit should be deployed. There are two schedulers: round robin and segregate scheduler.

The default one is round robin, this page describes what the segregate scheduler does and how to enable it.

How it works

Segregate scheduler is a scheduler that segregates the units between nodes by team.

First, what you need to do is to define a relation between a pool and teams. After that you need to register nodes with the pool metadata information, indicating to which pool the node belongs.

When deploying an application, the scheduler will choose among the nodes with the pool metadata information associated to the team owning the application being deployed.

Configuration and setup

To use the segregate scheduler you need to enable the segregate mode in tsuru.conf:

docker:
  segregate: true

Adding a pool

Using tsuru-admin you create a pool:

$ tsuru-admin docker-pool-add pool1

Adding teams to a pool

You can add one or more teams at once.

$ tsuru-admin docker-pool-teams-add pool1 team1 team2

$ tsuru-admin docker-pool-teams-add pool2 team3

Listing a pool

To list pools you do:

$ tsuru-admin docker-pool-list
+-------+-------------+
| Pools | Teams       |
+-------+-------------+
| pool1 | team1 team2 |
| pool2 | team3       |
+-------+-------------+

Registering a node with pool metadata

You can use the tsuru-admin with docker-node-add to register or create nodes with the pool metadata:

$ tsuru-admin docker-node-add --register address=http://localhost:4243 pool=pool1

Removing a pool

To remove a pool you do:

$ tsuru-admin docker-pool-remove pool1

Removing teams from a pool

You can remove one or more teams at once.

$ tsuru-admin docker-pool-teams-remove pool1 team1

$ tsuru-admin docker-pool-teams-remove pool1 team1 team2 team3