kubeadm-upgrade.md: add more details for upgrade rollbacks#18424
kubeadm-upgrade.md: add more details for upgrade rollbacks#18424k8s-ci-robot merged 1 commit intokubernetes:masterfrom
Conversation
b5143f0 to
9b31bd1
Compare
fabriziopandini
left a comment
There was a problem hiding this comment.
@neolit123 thanks
Only one curiosity, but nothing blocking
|
|
||
| `kubeadm-backup-etcd` contains a backup of the local etcd member data for this control-plane Node. | ||
| In case of an etcd upgrade failure and if the automatic rollback does not work, the contents of this folder | ||
| can be manually restored in `/var/lib/etcd`. In case of HA setup with multiple local etcd members |
There was a problem hiding this comment.
Are we skipping the etcd backed in case of multi-controlplane... I don't remember this detail 😳
There was a problem hiding this comment.
before writing this update, digging the source code i found that we clean the etcd local member backup in case of external etcd or HA:
https://github.com/kubernetes/kubernetes/blob/ce2102f3637134519ec189f096da5277af6072a6/cmd/kubeadm/app/phases/upgrade/staticpods.go#L595
In case of HA setup with multiple local etcd members
perhaps i need to clarify that the backup is always written, but after upgrade:
- the folder contents will be cleaned for HA setup.
- the folder contents will persist for single control-plane setup.
or maybe we need to change this logic and always keep the backup.
There was a problem hiding this comment.
I think that making consistent how we managed stacked etcd (HA or not) makes more sense
There was a problem hiding this comment.
this has to be a k/k PR.
in my opinion we should always persist the etcd backup for stacked.
i'm inclined to remove the last sentence from this k/website PR:
In case of HA setup with multiple local etcd members ...
WDYT?
There was a problem hiding this comment.
ok to remove the last sentence and open an issue (a PR) to track the change on update
There was a problem hiding this comment.
@fabriziopandini
i've updated the PR to only have the "external etcd" part of the sentence.
created kubernetes/kubeadm#1998 to track the k/k cleanup/
There was a problem hiding this comment.
/hold
until the discussion in kubernetes/kubeadm#1998 is resolved.
Introduce a couple of paragraphs explaining what backup folders kubeadm creates during upgrade.
9b31bd1 to
70a48b7
Compare
|
I'm going to this so that a future /lgtm and /hold cancel will let it go live. |
|
Alternatively, /close it it the decision goes another way. |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: sftim The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
/hold cancel the following PR is merging so that we now have the update in this current docs PR accurate: |
|
/lgtm |
…s#18424) Introduce a couple of paragraphs explaining what backup folders kubeadm creates during upgrade.
Introduce a couple of paragraphs explaining what backup folders kubeadm creates during upgrade.
note: this is not a new feature, thus i'm sending it to the
masterbranch here.fixes kubernetes/kubeadm#1694
/assign @fabriziopandini @jimangel
/sig cluster-lifecycle
/kind cleanup