Scribd
Upload
3 views

1.1 AutoScaling-in-Kubernetes

Uploaded by

anghaprapul
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

1.1 AutoScaling-in-Kubernetes

Uploaded by

anghaprapul
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

Kubernetes: Auto-Scaling

KUBERNETES : Advance of Kuebernetes

➤ AutoScale: Scaling the System to serve the requirement.


➤ Understand AutoScale: WaterTub Problem
➤ We want to ensure that once the first tub is 80% full, another
tub should be placed and the water should be sent to the
second tub.
KUBERNETES : Advance of Kuebernetes

➤ Things to Understand by Above Problem.


➤ Tub - the unit of scaling (What to scale?)
➤ 80% mark - metric and trigger for scaling (When to scale?)
➤ Pipe - the operation which enables scaling in this case (How
to scale?)
KUBERNETES : Advance of Kuebernetes

What to Scale?
➤ Pods: For a given application let's say you are running X
replicas, In case of more Traffic Y number of Pods can also be
executed on Cluster.
➤ For this our Nodes have enough resources available.
➤ Nodes: Capacity of all nodes put together represents your
cluster's capacity. If the workload demand goes beyond this
capacity, then you would have to add nodes to the cluster and
make sure the workload can be scheduled and executed
effectively.
KUBERNETES : Advance of Kuebernetes

When to Scale?
➤ By measuring a certain metric continuously and when the
metric crosses a threshold value, then acting on it by scaling a
certain resource.For this our Nodes have enough resources
available.
➤ For example, you might want to measure the average CPU
consumption of your pods and then trigger a scale operation
if the CPU consumption crosses 80%.
➤ For memory intensive applications, memory consumption
might be that metric.
KUBERNETES : Advance of Kuebernetes

How to Scale?
➤ CPU Based Scaling (HPA): Horizontal Pod Scaling
➤ With Horizontal Pod Autoscaling, Kubernetes automatically scales
the number of pods in a replication controller, deployment or
replica set based on observed CPU utilisation.
➤ Mostly this type of Scaling is based on CPU and Memory metrics.
➤ HPA is not applicable for DaemonSets.
➤ User can define the threshold and minimum and maximum scale to
which the deployment should scale.
➤ Create HPA for Deployment:
kubectl autoscale deployment <deployment-name> --min=2 --
max=5 --cpu-percent=80
KUBERNETES : Advance of Kuebernetes
How to Scale?
➤ Cluster AutoScaler
➤ Cluster autoscaler is used in Kubernetes to scale cluster i.e. nodes
dynamically.
➤ CA watches the pods continuously and if it finds that a pod cannot be
scheduled, then based on the PodCondition, it chooses to scale up.
➤ The cluster autoscaler requests a newly provisioned node if:
There are pending pods due to not having enough available cluster resources
to meet their requests and
The cluster or node pool has not reached the user-defined maximum node
count.
➤ Kubernetes detects the new node once it is provision by the underlying
infrastructure.
➤ The Kubernetes scheduler allocates the pending pods to the new node.
KUBERNETES : Advance of Kuebernetes

➤ To enable Metrics on Cluster, Cluster has to be started with


env variable:
ENABLE_CUSTOM_METRICS=TRUE
➤ An Example:
➤ Run Deployment with CPU request 200m.
➤ 200m = 200 miliCPUs or 20% core of Running Node. If Node
is 2 core then it’s still 20% of Single Core.
➤ User can Introduce AutoScaling at 50% of CPU uses(Which is
100m/10% of Code with the Pod)
Will see you in Next Lecture…

See you in next lecture …

You might also like