Kubernetes : Tweaking Your Settings

How to fine tune your settings to get the most out of OpenSquiggly.


After you’ve done the initial installation of OpenSquiggly on your Kubernetes cluster, you may want to adjust your settings after-the-fact to add functionality or change some important configuration values.

General Approach for Updating Settings

If you’ve installed OpenSquiggly using the provided Helm charts, then you should always make changes through Helm. You shouldn’t be using kubectl to go behind the scenes to make changes that were initially set by Helm. By always using Helm to make changes, you’ll ensure that the Helm history, upgrade, and rollback functions are in sync with the actual state of the cluster.

To change any Helm chart setting after the initial installation, use the command of the format:

helm upgrade your-release-name opensquiggly/allinone --reuse-values --set new-values-here

The --reuse-values flag will keep all existing values in the Helm installation unmodified except the ones specified after the --set flag.

For example, if your release is named opensquiggly1 and you want to change the dnsHostName to new.yourcompany.com while keeping all other settings unmodified do:

helm upgrade opensquiggly1 opensquiggly/allinone --reuse-values --set dnsHostName=new.yourcompany.com

Stopping & Restarting OpenSquiggly

Most of the time you won’t need to stop the OpenSquiggly instance (which in Kubernetes parlance would mean killing the OpenSquiggly pod), but occassionaly you may need to. One example of when you might need to stop the instance is if you are resizing the disk volume and are using a cloud provider that doesn’t support online resizing. Another example would be if you wanted to relocate the database from the default internal MongoDB database to an externally hosted database.

The start flag in the Helm chart can be used to set whether the pod is active or inactive.

To stop the OpenSquiggly instance, do:

helm upgrade release-name opensquiggly/allinone --reuse-values --set start=false

Afterwards, confirm that the OpenSquiggly pod is no longer running with the command:

kubectl get pods

The associated OpenSquiggly pod should no longer appear in the list of pods.

To restart the instance after you’ve stopped it, do:

helm upgrade release-name opensquiggly/allinone --reuse-values --set start=true

Aftwards, confirm the pod has restarted with the command:

kubectl get pods --watch

It may take a few minutes for the pod to fully initialize.

Resizing Volumes

We’ve created our Helm charts to make it easy to resize your disk space after the installation. By doing it this way, you don’t need to over allocate your disk space, and over pay for your hosting costs.

We recommend starting with a modest disk size in the order of 10 - 50 GB. Later, after you’ve attached a lot of mount points and run low on disk space, you can increase the disk size as-needed.

Note that you can never make the disk size smaller after you’ve increased it. This phenomenon also means you won’t be able to do a helm rollback to rollback to a prior revision that existed before resizing.

If Your Cloud Provider Supports Online Resizing

In many cases, depending on your cloud provider and the storage class you are using, your cluster may support online resizing of the disk space. This means you can resize the disk while keeping the OpenSquiggly instance running.

In this case, increasing the disk space is a simple matter of changing the diskSize property with the command:

helm upgrade release-name opensquiggly/allinone --reuse-values --set diskSize=new-size-in-gigabytes

For example, to change your disk space for the opensquiggly1 release to 50GB, do:

helm upgrade opensquiggly1 opensquiggly/allinone --reuse-values --set diskSize=50

If Your Cloud Provider Requires Offline Resizing

If your cloud provider doesn’t support online resizing, then you’ll need to stop the instance, resize the disk, and restart.

Here is the sequence of commands for such:

# Stop the instance
helm upgrade release-name opensquiggly/allinone --reuse-values --set start=false

# Wait for pod to shut down (should shut down immediately)
kubectl get pods --watch

# Do the resize independent of the restart. This gives the new size a chance to
# take affect before restarting
helm upgrade release-name opensquiggly/allinone --reuse-values --set diskSize=new-size

# Restart only after resizing
helm upgrade release-name opensquiggly/allinone --reuse-values --set start=true

# Wait for pod to start
kubectl get pods --watch

Revision History

Each time you make any changes to a Helm release, a new revision is created with the updated settings. To view the list of revisions do:

helm history release-name-here

Viewing Current Values

To see the list of all Helm chart values for the current revision, do:

helm get values release-name-here --all

Viewing Values for a Prior Revision

To see the list of all Helm chart values for a past revision, do:

helm get values release-name-here --all --revision rev-number-here


To perform a rollback to a prior revision of the release, do:

helm rollback release-name-here rev-number-here

This creates a new revision with the settings equal to that of the specified prior revision.