Downgrade updates after testing #1544
Conversation
tmjd
force-pushed
the
fixup-downgrade
branch
from
48d6343 to
cbca4f0
Compare
| before downgrading the {{site.prodname}} components. | ||
| 1. Follow the steps in | ||
| [Aborting the upgrade](#aborting-the-upgrade-when-using-the-etcd-datastore) | ||
| can be done before downgrading the {{site.prodname}} components. |
There was a problem hiding this comment.
During my test, I found reuse of Aborting the upgrade section could cause confusion because user is not actually aborting upgrade. Maybe something like Run `calico-upgrade abort` to instruct calico to use previous version datastore.
| _Docs for this coming soon!_ | ||
|
|
||
| ## Aborting the upgrade | ||
| ## Aborting the upgrade when using the etcd datastore | ||
|
|
||
| If you have not upgraded the {{site.prodname}} components then follow |
There was a problem hiding this comment.
This is a little confusing after being directed here by the Downgrade instructions. I'll rework or remove it.
| > before downgrading will lose networking. | ||
| {: .alert .alert-danger} | ||
|
|
||
| - Follow the steps to [Abort the upgrade when using the etcd |
There was a problem hiding this comment.
The abort procedure is discussed in the migrate section so I don't think it belongs here, I think we can delete this.
There was a problem hiding this comment.
Is what you are talking about in a PR still? In master I don't see an abort procedure, the only abort mention is a reference to it in the Errors note.
tmjd
force-pushed
the
fixup-downgrade
branch
from
f241056 to
3e4ba1a
Compare
| - `bgpconfigurations.crd.projectcalico.org` | ||
| - `felixconfigurations.crd.projectcalico.org` | ||
| - `clusterinformations.crd.projectcalico.org` |
There was a problem hiding this comment.
what about networkpolicies.crd.projectcalico.org?
| to try again. | ||
|
|
||
| ## Deleting Calico data from the Kubernetes API datastore after a downgrade |
There was a problem hiding this comment.
Need to add a bullet around line 17 explaining when you might need to execute this procedure and linking to the anchor link here (like the other bullets)
| - `felixconfigurations.crd.projectcalico.org` | ||
| - `clusterinformations.crd.projectcalico.org` | ||
|
|
There was a problem hiding this comment.
I would put the expected output below the command that returns it
tmjd
force-pushed
the
fixup-downgrade
branch
from
3e4ba1a to
c3ea326
Compare
| @@ -14,6 +14,9 @@ following conditions. | |||
|
|
|||
| - [A data migration attempt failed partway through, leaving the etcdv3 datastore | |||
| with some, but not all of your etcvd2 data](#deleting-calico-data-from-etcdv3-after-a-partial-migration). | |||
|
|
|||
| - [You upgraded to Calico v3.0 but then downgraded to your previous version | |||
There was a problem hiding this comment.
Since they upgraded from v2.6.4, then we know their previous version and can change "your previous version" to "v2.6.4"
There was a problem hiding this comment.
I think you should specify in this bullet text that this is for the "Kubernetes API datastore" (versus etcd)
| @@ -14,6 +14,9 @@ following conditions. | |||
|
|
|||
| - [A data migration attempt failed partway through, leaving the etcdv3 datastore | |||
| with some, but not all of your etcvd2 data](#deleting-calico-data-from-etcdv3-after-a-partial-migration). | |||
|
|
|||
| - [You upgraded to Calico v3.0 but then downgraded to your previous version | |||
| but need to clean up before upgrading again](#deleting-calico-data-from-the-kubernetes-api-datastore-after-a-downgrade). | |||
There was a problem hiding this comment.
I would change "but" to "and" here, I think it's more correct
- Add steps to delete v3.0 KDD CRDs after downgrade
tmjd
force-pushed
the
fixup-downgrade
branch
from
c3ea326 to
7d3055a
Compare
I tried to model this a little after some of the other upgrade docs.
Release Note