openstack-helm/doc/source/devref/pod-disruption-budgets.rst
Tin Lam fdfca4b0fb Fix handful of trivial items in docs
This fixes up typographical errors, capitalization issues and minor
stylistic formatting in the documentations.

Change-Id: I1951a90d48cc22e7f8c0516fc53f4018ef0a9a5d
Signed-off-by: Tin Lam <tin@irrational.io>
2018-05-25 19:26:03 -05:00

1.3 KiB

Pod Disruption Budgets

OpenStack-Helm leverages PodDistruptionBudgets to enforce quotas that ensure that a certain number of replicas of a pod are available at any given time. This is particularly important in the case when a Kubernetes node needs to be drained.

These quotas are configurable by modifying the minAvailable field within each PodDistruptionBudget manifest, which is conveniently mapped to a templated variable inside the values.yaml file. The min_available within each service's values.yaml file can be represented by either a whole number, such as 1, or a percentage, such as 80%. For example, when deploying 5 replicas of a pod (such as keystone-api), using min_available: 3 would enforce policy to ensure at least 3 replicas were running, whereas using min_available: 80% would ensure that 4 replicas of that pod are running.

Note: The values defined in a PodDisruptionBudget may conflict with other values that have been provided if an operator chooses to leverage Rolling Updates for deployments. In the case where an operator defines a maxUnavailable and maxSurge within an update strategy that is higher than a minAvailable within a pod disruption budget, a scenario may occur where pods fail to be evicted from a deployment.