Triage is an important part of maintaining the health of the minikube repo.
A well organized repo allows maintainers to prioritize feature requests, fix bugs, and respond to users facing difficulty with the tool as quickly as possible.
Triage includes:
- Labeling issues
- Responding to issues
- Closing issues
If you're interested in helping out with minikube triage, this doc covers the basics of doing so.
Additionally, if you'd be interested in participating in our weekly triage meeting, please fill out this [form](https://forms.gle/vNtWZSWXqeYaaNbU9) to express interest. Thank you!
# Daily Triage
Daily triage has two goals:
1. Responsiveness for new issues
1. Responsiveness when explicitly requested information was provided
The list of outstanding items are at http://tinyurl.com/mk-tparty/daily-triage - it covers:
1. Issues without a `kind/` or `triage/` label
1. Issues without a `priority/` label
1.`triage/needs-information` issues which the user has followed up on, and now require a response.
## Categorization
The most important level of categorizing the issue is defining what type it is.
We typically want at least one of the following labels on every issue, and some issues may fall into multiple categories:
-`triage/support` - The default for most incoming issues
-`kind/bug` - When it’s a bug or we aren’t delivering the best user experience
Other possibilities:
-`kind/feature`- Identify new feature requests
-`kind/flake` - Used for flaky integration or unit tests
-`kind/cleanup` - Cleaning up/refactoring the codebase
-`kind/documentation` - Updates or additions to minikube documentation
-`kind/ux` - Issues that involve improving user experience
-`kind/security` - When there's a security vulnerability in minikube
If the issue is specific to an operating system, hypervisor, container, addon, or Kubernetes component:
**os/[operating system]** - When the issue appears specific to an operating system
-`os/linux`
-`os/macos`
-`os/windows`
**co/[driver]** - When the issue appears specific to a driver
-`co/hyperkit`
-`co/hyperv`
-`co/kvm2`
-`co/none-driver`
-`co/docker-driver`
-`co/virtualbox`
**co/[kubernetes component]** - When the issue appears specific to a k8s component
-`co/apiserver`
-`co/etcd`
-`co/coredns`
-`co/dashboard`
-`co/kube-proxy`
-`co/kubeadm`
-`co/kubelet`
-`co/kubeconfig`
Other useful tags:
Did an **Event** occur that we can dedup similar issues against?
-`ev/CrashLoopBackoff`
-`ev/Panic`
-`ev/Pending`
-`ev/kubeadm-exit-1`
Suspected **Root cause**:
-`cause/vm-environment`
-`cause/invalid-kubelet-options`
**Help wanted?**
`Good First Issue` - bug has a proposed solution, can be implemented w/o further discussion.
`Help wanted` - if the bug could use help from a contributor
## Prioritization
If the issue is not `triage/support`, it needs a [priority label](https://github.com/kubernetes/community/blob/master/contributors/guide/issue-triage.md#define-priority):
`priority/critical-urgent` - someones top priority ASAP, such as security issue, user-visible bug, or build breakage. Rarely used.
`priority/important-soon`: in time for the next two releases (8 weeks)
`priority/important-longterm`: 2-4 releases from now
`priority/backlog`: agreed that this would be good to have, but no one is available at the moment. Consider tagging as `help wanted`
`priority/awaiting-more-evidence`: may be more useful, but there is not yet enough support.
# Weekly Triage
Weekly triage has three goals:
1. Catching up on unresponded issues
1. Reviewing and closing PR’s
1. Closing stale issues
The list of outstanding items can be found at http://tinyurl.com/mk-tparty/weekly-triage.
## Post-Release Triage
Post-release triage occurs after a major release (around every 4-6 weeks).
It focuses on:
1. Closing bugs that have been resolved by the release
1. Reprioritizing bugs that have not been resolved by the release
1. Letting users know if we believe that there is still an issue
This includes reviewing:
1. Every issue that hasn’t been touched in the last 2 days
1. Re-evaluation of long-term issues
1. Re-evaluation of short-term issues
## Responding to Issues
### Needs More Information
A sample response to ask for more info:
> I don’t yet have a clear way to replicate this issue. Do you mind adding some additional details. Here is additional information that would be helpful:
>
> \* The exact `minikube start` command line used
>
> \* The full output of the `minikube start` command, preferably with `--alsologtostderr -v=4` for extra logging.
>
> \* The full output of `minikube logs`
>
> \* The full output of `kubectl get po -A`
>
>
>
> Thank you for sharing your experience!
Then: Label with `triage/needs-information`.
### Issue might be resolved
If you think a release may have resolved an issue, ask the author to see if their issue has been resolved:
> Could you please check to see if minikube <x> addresses this issue? We've made some changes with how this is handled, and improved the minikube logs output to help us debug tricky cases like this.
Then: Label with `triage/needs-information`.
## Closing with Care
Issues typically need to be closed for the following reasons:
- The issue has been addressed
- The issue is a duplicate of an existing issue
- There has been a lack of information over a long period of time
In any of these situations, we aim to be kind when closing the issue, and offer the author action items should they need to reopen their issue or still require a solution.
Samples responses for these situations include:
### Issue has been addressed
>@author: I believe this issue is now addressed by minikube v1.4, as it <reason>. If you still see this issue with minikube v1.4 or higher, please reopen this issue by commenting with `/reopen`
>
>Thank you for reporting this issue!
Then: Close the issue
### Duplicate Issue
>This issue appears to be a duplicate of #X, do you mind if we move the conversation there?
>
>This way we can centralize the content relating to the issue. If you feel that this issue is not in fact a duplicate, please re-open it using `/reopen`. If you have additional information to share, please add it to the new issue.
>
>Thank you for reporting this!
Then: Label with `triage/duplicate` and close the issue.
### Lack of Information
If an issue hasn't been active for more than four weeks, and the author has been pinged at least once, then the issue can be closed.
>Hey @author -- hopefully it's OK if I close this - there wasn't enough information to make it actionable, and some time has already passed. If you are able to provide additional details, you may reopen it at any point by adding /reopen to your comment.
>
>Here is additional information that may be helpful to us:
>
>\* Whether the issue occurs with the latest minikube release
>
>\* The exact `minikube start` command line used
>
>\* The full output of the `minikube start` command, preferably with `--alsologtostderr -v=3` for extra logging.
