Automatically grow PVCs in OCP IBM Cloud

Summary and Key concepts

Summary:

This article explains how to use Portworx Autopilot to automatically expand PersistentVolumeClaims (PVCs) when they begin to run out of space. Autopilot monitors PVC usage, and when a threshold is reached, such as exceeding 50% usage, it automatically resizes the PVC. The example provided showcases an Autopilot rule that monitors PostgreSQL PVCs and doubles their size when usage exceeds 50%, up to a maximum size of 400GiB. The article also provides steps to create the necessary Kubernetes resources (namespaces, storage class, PVCs, and PostgreSQL application) and monitor the progress of Autopilot actions using Kubernetes events.

Kubernetes Concepts:

• PersistentVolumeClaim (PVC): A request for storage by Kubernetes users. Autopilot automatically resizes PVCs based on predefined conditions.
• StorageClass: Defines how storage is provisioned in a Kubernetes cluster. The article includes an example of a Portworx-backed storage class.
• Namespace: Used to organize Kubernetes objects. In the example, the namespaceSelector ensures the rule only applies to specific namespaces.

Portworx Concepts:

• Autopilot: Automates storage management tasks such as resizing PVCs when they reach a certain usage threshold.
• AutopilotRule: A custom resource in Portworx that defines the conditions to monitor and the actions to take, such as PVC resizing.

Using Autopilot to Autogrow PVCs

You can use Autopilot to expand PVCs automatically when they begin to run out of space. Autopilot monitors the metrics in your cluster (e.g., via Prometheus) and detects high usage conditions. Once high usage conditions occur, Autopilot talks with your cluster to resize the PVC.

An AutopilotRule that has 4 main parts:

1. PVC Selector Matches labels on the PVCs.
2. Namespace Selector Matches labels on the Kubernetes namespaces the rule should monitor. This is optional, and the default is all namespaces.
3. Metric conditions on the PVC to monitor.
4. PVC resize action to perform once the metric conditions are met.

The following example section shows the actual YAML for this.

Example

The following example Autopilot rule expands Postgres PVCs by 100% whenever their usage exceeds 50% up to a maximum size of 400GiB:

apiVersion: autopilot.libopenstorage.org/v1alpha1
kind: AutopilotRule
metadata:
 name: volume-resize
spec:
  ##### selector filters the objects affected by this rule given labels
  selector:
    matchLabels:
      app: postgres
  ##### namespaceSelector selects the namespaces of the objects affected by this rule
  namespaceSelector:
    matchLabels:
      type: db
  ##### conditions are the symptoms to evaluate. All conditions are AND'ed
  conditions:
    # volume usage should be less than 50%
    expressions:
    - key: "100 * (px_volume_usage_bytes / px_volume_capacity_bytes)"
      operator: Gt
      values:
        - "50"
  ##### action to perform when condition is true
  actions:
  - name: openstorage.io.action.volume/resize
    params:
      # resize volume by scalepercentage of current size
      scalepercentage: "100"
      # volume capacity should not exceed 400GiB
      maxsize: "400Gi"

Consider the key sections in this spec.

• selector and namespaceSelector
• conditions
• action

The selector determines what objects are acted on by the Autopilot rule by looking for PVCs with the app: postgres label. Similarly, the namespaceSelector filters PVCs by namespaces and only includes PVCs from namespaces that contain the type: db label. Hence, this rule applies only to PVCs running Postgres in the DB namespaces.

note

You can also use matchExpressions with selector and namespaceSelector. For more information, refer to selector and namespaceSelector.

The conditions section determines the threshold criteria dictating when the rule has to perform its action. In this example, that criteria has the following formula:

100 * (px_volume_usage_bytes / px_volume_capacity_bytes) gives the volume usage percentage and the Gt operator puts a condition that volume usage percentage has exceeded 50%.

Conditions are combined using AND logic, requiring all conditions to be true for the rule to trigger.

The actions section specifies what action Portworx performs when the conditions are met. Action parameters modify action behavior, and different actions contain different action parameters.

Perform the following steps to deploy this example:

Create specs

Application and PVC specs

First, create the storage and application spec files:

1. Create namespaces.yaml and place the following content inside it.

   apiVersion: v1
   kind: Namespace
   metadata:
     name: pg1
     labels:
       type: db
   ---
   apiVersion: v1
   kind: Namespace
   metadata:
     name: pg2
     labels:
       type: db

2. Create postgres-sc.yaml and place the following content inside it.

   ##### Portworx storage class
   apiVersion: storage.k8s.io/v1
   kind: StorageClass
   metadata:
     name: postgres-pgbench-sc
   provisioner: pxd.portworx.com
   parameters:
     repl: "2"
   allowVolumeExpansion: true

3. Create postgres-vol.yaml and place the following content inside it.

   kind: PersistentVolumeClaim
   apiVersion: v1
   metadata:
     name: pgbench-data
     labels:
       app: postgres
   spec:
     storageClassName: postgres-pgbench-sc
     accessModes:
       - ReadWriteOnce
     resources:
       requests:
         storage: 10Gi
   ---
   kind: PersistentVolumeClaim
   apiVersion: v1
   metadata:
     name: pgbench-state
   spec:
     storageClassName: postgres-pgbench-sc
     accessModes:
       - ReadWriteOnce
     resources:
       requests:
         storage: 1Gi

4. Createpostgres-app.yaml and place the following content inside it.

   The application in this example is a PostgreSQL database with a pgbench sidecar. The SIZE environment variable in this spec instructs pgbench to write 70GiB of data to the volume. Since the PVC is only 10GiB in size, Autopilot must resize the PVC when needed.

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: pgbench
     labels:
       app: pgbench
   spec:
     selector:
       matchLabels:
         app: pgbench
     strategy:
       rollingUpdate:
         maxSurge: 1
         maxUnavailable: 1
       type: RollingUpdate
     replicas: 1
     template:
       metadata:
         labels:
           app: pgbench
       spec:
         schedulerName: stork
         containers:
           - image: postgres:9.5
             name: postgres
             ports:
             - containerPort: 5432
             env:
             - name: POSTGRES_USER
               value: pgbench
             - name: POSTGRES_PASSWORD
               value: superpostgres
             - name: PGBENCH_PASSWORD
               value: superpostgres
             - name: PGDATA
               value: /var/lib/postgresql/data/pgdata
             volumeMounts:
             - mountPath: /var/lib/postgresql/data
               name: pgbenchdb
           - name: pgbench
             image: portworx/torpedo-pgbench:latest
             imagePullPolicy: "Always"
             env:
               - name: PG_HOST
                 value: 127.0.0.1
               - name: PG_USER
                 value: pgbench
               - name: SIZE
                 value: "70"
             volumeMounts:
             - mountPath: /var/lib/postgresql/data
               name: pgbenchdb
             - mountPath: /pgbench
               name: pgbenchstate
         volumes:
         - name: pgbenchdb
           persistentVolumeClaim:
             claimName: pgbench-data
         - name: pgbenchstate
           persistentVolumeClaim:
             claimName: pgbench-state

AutopilotRule spec

Once you've created your storage and application specs, you can create an AutopilotRule that controls them.

important

It is recommended to set up only one Autopilot rule for each PVC, because Autopilot does not support more than one rule per PVC.

Create a YAML spec for the autopilot rule named autopilotrule-example.yaml and place the following content inside it:

apiVersion: autopilot.libopenstorage.org/v1alpha1
kind: AutopilotRule
metadata:
  name: volume-resize
spec:
  ##### selector filters the objects affected by this rule given labels
  selector:
    matchLabels:
      app: postgres
  ##### namespaceSelector selects the namespaces of the objects affected by this rule
  namespaceSelector:
    matchLabels:
      type: db
  ##### conditions are the symptoms to evaluate. All conditions are AND'ed
  conditions:
    # volume usage should be less than 50%
    expressions:
    - key: "100 * (px_volume_usage_bytes / px_volume_capacity_bytes)"
      operator: Gt
      values:
        - "50"
  ##### action to perform when condition is true
  actions:
  - name: openstorage.io.action.volume/resize
    params:
      # resize volume by scalepercentage of current size
      scalepercentage: "100"
      # volume capacity should not exceed 400GiB
      maxsize: "400Gi"

Apply specs

Once you've designed your specs, deploy them.

oc apply -f autopilotrule-example.yaml
oc apply -f namespaces.yaml
oc apply -f postgres-sc.yaml
oc apply -f postgres-vol.yaml -n pg1
oc apply -f postgres-vol.yaml -n pg2
oc apply -f postgres-app.yaml -n pg1
oc apply -f postgres-app.yaml -n pg2

Monitor

Notice that the pgbench pods in the pg1 and pg2 namespace will start filling up the pgbench-data PVCs. As the PVC usage starts exceeding 50%, Autopilot will resize the PVCs.

You can use the following command to get all the events generated for the volume-resize rule:

oc get events --field-selector involvedObject.kind=AutopilotRule,involvedObject.name=volume-resize --all-namespaces --sort-by .lastTimestamp

NAMESPACE   LAST SEEN   TYPE     REASON       KIND            MESSAGE
default     21m         Normal   Transition   AutopilotRule   rule: pvc-xxxxxxxx-xxxx-xxxx-xxxx-aa931955114b transition from Initializing => Normal
default     21m         Normal   Transition   AutopilotRule   rule: pvc-xxxxxxxx-xxxx-xxxx-xxxx-aa931955114b transition from Initializing => Normal
default     9m52s       Normal   Transition   AutopilotRule   rule: pvc-xxxxxxxx-xxxx-xxxx-xxxx-aa931955114b transition from Initializing => Normal
default     9m48s       Normal   Transition   AutopilotRule   rule: pvc-xxxxxxxx-xxxx-xxxx-xxxx-aa931955114b transition from Initializing => Normal