https://github.com/luciferinlove/aws-k8s-ebs-autoscaler
Enlarges the size of AWS EBS volumes directly or as Kubernetes PVC.
https://github.com/luciferinlove/aws-k8s-ebs-autoscaler
aws ebs ebs-snapshots ebs-volumes go golang kubernetes pvc
Last synced: 15 days ago
JSON representation
Enlarges the size of AWS EBS volumes directly or as Kubernetes PVC.
- Host: GitHub
- URL: https://github.com/luciferinlove/aws-k8s-ebs-autoscaler
- Owner: LuciferInLove
- License: apache-2.0
- Created: 2021-05-04T13:26:16.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2021-05-04T19:53:31.000Z (over 4 years ago)
- Last Synced: 2025-06-09T13:35:45.801Z (4 months ago)
- Topics: aws, ebs, ebs-snapshots, ebs-volumes, go, golang, kubernetes, pvc
- Language: Go
- Homepage:
- Size: 37.1 KB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE
Awesome Lists containing this project
README
[](https://goreportcard.com/report/github.com/LuciferInLove/aws-k8s-ebs-autoscaler)
# aws-k8s-ebs-autoscaler
Enlarges the size of AWS EBS volumes directly or as Kubernetes PVC. It can be used in pair with [alertmanager-webhook-receiver](https://github.com/LuciferInLove/alertmanager-webhook-receiver). You can see a sample of usage in the How it works section below.
NOTE: This is an alpha version, you shouldn't use it in a production environment.
NOTE: Any value of the percents flag causes the size to increase to 1 Gb minimum. For example, if the size of the EBS volume is 5 Gb and the percents is less than 35, the EBS volume size will be increased to 1 Gb.
## Usage
```
Usage of aws-k8s-ebs-autoscaler:
-dry-run
If true, only show the result without enlarging the volume. (default false)
-k8s-snapshot-class string
The name of the VolumeSnapshotClass resource, which is used to create snapshots in Kubernetes. (default "csi-aws-vsc")
-log-level string
Only log messages with the given severity or above. One of: [debug, info, warn, error] (default "info")
-mount-point string
Mount point of the volume to be enlarged. (required if pvc isn't set)
-percents int
By what percentage to increase. (default 20)
-proc-path string
procfs mountpoint. (default "/proc")
-pvc string
PVC ID of the volume to be enlarged. (required if mount-point isn't set)
-pvc-namespace string
Kubernetes namespace where pvc is located. (required if mount-point isn't set)
-snapshot
If true, create a volume snapshot. (default false)
-sys-path string
sysfs mountpoint. (default "/sys")
-wait-for-modifying
If true, wait for enlarging the volume to be completed. (default false)
```
**aws-k8s-ebs-autoscaler** reads namespace from `/var/run/secrets/kubernetes.io/serviceaccount/namespace` if namespaces aren't defined in command line flags. It also reads [Kubernetes Service Account Token](https://kubernetes.io/docs/reference/access-authn-authz/authentication/#service-account-tokens) from `/var/run/secrets/kubernetes.io/serviceaccount/token`.
## How it works
**aws-k8s-ebs-autoscaler** was created to run mainly as Kubernetes Job in pair with alertmanager-webhook-receiver. alertmanager-webhook-receiver responds to monitoring events and triggers Kubernetes Jobs with labels as parameters. For example, the event of insufficient disk space appears as an alert in the alertmanager. Alertmanager sends webhook to alertmanager-webhook-receiver. alertmanager-webhook-receiver takes the mount point or PVC from the labels and runs **aws-k8s-ebs-autoscaler** as Kubernetes Job.
Note that false and multiple consecutive alerts are the responsibility of the monitoring system, not of the alertmanager-webhook-receiver or **aws-k8s-ebs-autoscaler**.
**aws-k8s-ebs-autoscaler** performs actions depending on what was passed as an argument, mount-point or pvc.
### If mount-point is received in arguments
NOTE: Linux file system won't automatically extend after the volume enlargement. You could run **aws-k8s-ebs-autoscaler** as an init container and then run a container with utilities to extend the Linux file system, but it's better to use external tools for security reasons. Or you can use such tools as [embiggen-disk](https://github.com/bradfitz/embiggen-disk). Read [this](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/recognize-expanded-volume-linux.html) doc.
* **aws-k8s-ebs-autoscaler** searches for volume serial number by the mount point. In the case of EBS the serial number is EBS VolumeID.
* If the snapshot flag was provided as true, it creates an EBS volume snapshot.
* If the dry-run flag was provided as true, **aws-k8s-ebs-autoscaler** only shows information about enlarging.
* If not, it enlarges the EBS volume by a percentage, defined in the percents flag.
* If the wait-for-modifying flag was provided as true, **aws-k8s-ebs-autoscaler** waits for the in-use status of the EBS volume.
### If pvc is received in arguments
NOTE: The allowVolumeExpansion and ExpandInUsePersistentVolumes options should be enabled in your Kubernetes cluster for the PVC auto enlarging. Read [this](https://kubernetes.io/blog/2018/07/12/resizing-persistent-volumes-using-kubernetes/) doc.
* **aws-k8s-ebs-autoscaler** gets the PVC metadata in the given Kubernetes namespace provided in the pvc-namespace flag.
* If the snapshot flag was provided as true, it creates a [Kubernetes VolumeSnapshot](https://kubernetes.io/docs/concepts/storage/volume-snapshots/) resource. Note about the k8s-snapshot-class flag.
* If the dry-run flag was provided as true, **aws-k8s-ebs-autoscaler** only shows information about enlarging.
* If not, it enlarges the PVC size.
* If the wait-for-modifying flag was provided as true, **aws-k8s-ebs-autoscaler** waits for the Ready status of the PVC.