Skip to content

Conversation

@emosbaugh
Copy link
Member

No description provided.

@emosbaugh emosbaugh requested a review from a team as a code owner October 4, 2024 21:30
@netlify
Copy link

netlify bot commented Oct 4, 2024

Deploy Preview for replicated-docs ready!

Name Link
🔨 Latest commit ba3820f
🔍 Latest deploy log https://app.netlify.com/sites/replicated-docs/deploys/67059def37297f00083d3dfd
😎 Deploy Preview https://deploy-preview-2723--replicated-docs.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify site configuration.

@netlify
Copy link

netlify bot commented Oct 4, 2024

Deploy Preview for replicated-docs-upgrade ready!

Name Link
🔨 Latest commit ba3820f
🔍 Latest deploy log https://app.netlify.com/sites/replicated-docs-upgrade/deploys/67059def534ccc00081a7bca
😎 Deploy Preview https://deploy-preview-2723--replicated-docs-upgrade.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify site configuration.

@replicated-ci replicated-ci added the type::docs Improvements or additions to documentation label Oct 4, 2024
@emosbaugh emosbaugh requested a review from ajp-io October 4, 2024 22:28
paigecalvert
paigecalvert previously approved these changes Oct 4, 2024
@ajp-io
Copy link
Member

ajp-io commented Oct 7, 2024

@paigecalvert We can talk tomorrow, but we can make a few changes here:

  • We moved all of our main data to /var/lib/embedded-cluster, so we don't really need to document /var/lib/embedded-cluster/k0s, /var/lib/embedded-cluster/openebs-local, or /tmp anymore. We can just mention /var/lib/embedded-cluster, which needs to be at least 40 Gi and less than 80% full.
  • I wonder if the --data-dir flag should be documented outside of the requirements section. Maybe requirements just say a 40 Gi installation directory, and mention that we use /var/lib/embedded-cluster by default, but the user can choose an alternative with --data-dir. Like what we do for the ports.

@emosbaugh
Copy link
Member Author

Additionally, we could add a limitation that if you use an alternate data-dir, you must use the same location when restoring.

It is not possible to change the port for the Admin Console during a restore with Embedded Cluster. For more information, see [Disaster Recovery for Embedded Cluster (Alpha)](/vendor/embedded-disaster-recovery).
It is not possible to change the port for the Admin Console during a restore with Embedded Cluster. For more information, see [Disaster Recovery for Embedded Cluster (Alpha)](/vendor/embedded-disaster-recovery).

## Change the Default Data Directory {#data-dir}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

^ Added "Change the Default Data Directory" section in the Embedded Cluster online install topic next just below the existing section of the changing the default kotsadm/LAM ports

```
Where `APP_SLUG` is the unique application slug.

### Limtiations
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

^ Added limitations related to changing the default dir, including that you have to use the same dir during a restore

* The filesystem at `/var/openebs` has 5Gi or more of total space
* The filesystem at `/tmp` has 5Gi or more of total space

* The filesystem at `/var/lib/embedded-cluster` has 40Gi or more of total space and must be less than 80% full
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

^ Updated the disk space requirements to clarify that you just need the embedded-cluster dir with 40gi+ of space and less than 80% full.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

With the change that all data is stored in the /var/lib/embedded-cluster dir, it seemed like this whole storage requirements partial could be removed. I'm thinking that if there's anything from this that we want to tell people (like some examples of what's stored in /var/lib/embedded-cluster), I could add that to the disk storage requirement

#### `/tmp`
- Temporary space during installation

#### Additional Directories and Files
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

^ I could pull this info out of the partial and add it back to the docs. Wasn't sure if we still wanted to list these out

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@paigecalvert I missed this comment before. The additional list of directories would be good to add back in, just so people have a place to see places where some data is written, even if it's a lot less than is written to our data directory.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Gotcha, will do

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Added in here: #2741

@paigecalvert
Copy link
Contributor

@emosbaugh @ajp-io this is ready for a review from you two. I think I addressed all the latest notes (and added comments to explain what I changed).

### Bug Fixes {#bug-fixes-1-11-0}

* Fixes several issues that caused node resets to fail. Single-node clusters are no longer drained before being reset. Resets will no longer fail with the error `unable to get installation` if the installation failed early on. And node resets will now work if bind mounts are used for `/var/lib/embedded-cluster`, `/var/lib/k0s`, and `/var/openebs`.
* Fixes several issues that caused node resets to fail. Single-node clusters are no longer drained before being reset. Resets will no longer fail with the error `unable to get installation` if the installation failed early on. And node resets will now work if bind mounts are used for `/var/lib/embedded-cluster`, `/var/lib/embedded-cluster/k0s`, and `/var/lib/embedded-cluster/openebs-local`.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we want to retroactively change release notes? The old paths were true for this release, and moving forward we're just not mentioning the k0s and openebs directories. I don't know what's best practice here.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I usually don't retroactively change release notes since they are a point-in-time representation. The exception would be if we need to pull a release or if we forgot to add something

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

So we should undo this, right?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Undid it and reapproved the PR 👍

paigecalvert and others added 2 commits October 8, 2024 11:28
Co-authored-by: Alex Parker <7272359+ajp-io@users.noreply.github.com>
@emosbaugh emosbaugh merged commit 43f02f1 into main Oct 10, 2024
5 checks passed
@emosbaugh emosbaugh deleted the emosbaugh/sc-108561/consolidate-so-that-every-possible-directory branch October 10, 2024 17:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

type::docs Improvements or additions to documentation type::feature

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants