rancher
/
fleet-docs
mirror of https://github.com/rancher/fleet-docs.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Add troubleshooting intro 

This provides an overview of where and when to gather clues about what
Fleet may be doing wrong, before the section explaining _how_ to do so.
This commit is contained in:
Corentin Néau 2024-09-20 17:39:39 +02:00
parent 06f75b4073
commit ff1e2f8823
No known key found for this signature in database
GPG Key ID: E92A451AA7F19EFF
1 changed files with 21 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
docs/troubleshooting.md
View File

@ -2,6 +2,27 @@
This section contains commands and tips to troubleshoot Fleet.
## Where to look for root causes of issues
The first things to check when Fleet behaves unexpectedly would be:
* `fleet-controller` logs on the management cluster: has Fleet failed to reconcile any resource's (bundle, bundle
deployment) current state with its expected state?
* `gitjob` pod logs on the management cluster: has Fleet encountered any issue while creating jobs to generate new
bundles for new commits found in monitored git repositories?
* status of the `GitRepo` for which resources are not properly deployed:
* How many `Ready Bundle Deployments` does it list?
* How many bundle deployments are listed as expected? How many do you expect to see?
* Keep in mind that a `GitRepo` creates a bundle per path; each bundle leads to as many bundle deployments
as there are target clusters. A mismatch between `Desired Ready Clusters` and your own expectation could point to a
targeting issue.
* Which resources are listed in the `GitRepo`'s status?
* Which commit appears in the `GitRepo`'s status?
If the issue is specific to a target cluster, one might want to check Fleet agent logs on that cluster: has Fleet failed
to deploy a bundle deployment on that cluster?
The next section explains how to run all these checks.
## **How Do I...**