nil pointer dereference when using a wrong selector #260
Assignees
Milestone
Comments
|
Thank you @prometherion |
gianlucam76
pushed a commit
to gianlucam76/event-manager
that referenced
this issue
Jan 11, 2024
gianlucam76
pushed a commit
to gianlucam76/sveltosctl
that referenced
this issue
Jan 11, 2024
gianlucam76
pushed a commit
to gianlucam76/event-manager
that referenced
this issue
Jan 11, 2024
gianlucam76
pushed a commit
to gianlucam76/addon-constraint-controller
that referenced
this issue
Jan 11, 2024
gianlucam76
pushed a commit
to gianlucam76/access-manager
that referenced
this issue
Jan 11, 2024
This was referenced Jan 11, 2024
Merged
AdamSadek
added a commit
to AdamSadek/sveltosctl
that referenced
this issue
Jun 5, 2024
Forgot to check for clusterType
Fixing name of func and adding clusterType to function arguments
removing instance as it will now use the passed in dc
experimenting
Making adjustments
importing apierrors & metav1
fixing managed dc
directly setting ClusterName and ClusterType?
Resolving syntax errors
removing comma
Replacing registry name in 'manifest/sveltosctl.yaml' and 'Make File' script, from gianlucam76 to projectsveltos.
Techsupport
Introduce CRD and reconciler.
Following is an example of techsupport instance that every hour
collects from each cluster matching cluster selector env=fv:
1. logs of pods in namespace kube-system, collecting the last 10 minutes
worth of logs (600 seconds);
2. all namespaces will be dumped as YAML.
```
apiVersion: utils.projectsveltos.io/v1alpha1
kind: Techsupport
metadata:
name: hourly
spec:
clusterSelector: env=fv
schedule: "00 * * * *"
storage: /techsupport
logs:
- namespace: kube-system
sinceSeconds: 600
resources:
- group: ""
version: v1
kind: Namespace
```
Prepare for release v0.6.0
Prepare dev to main sync
Prepare for release v0.7.0
Fix Makefile
Prepare dev to main merge
Prepare for release v0.7.1
Prepare dev to main merge
Release v0.7.2
Prepare dev to main merge
Advance pins
controller-tools: v0.11.3
controller-runtime: v0.14.5
cluster-api: v1.4.0-rc.1
k8s: v0.26.0
Advance pins
Collect EventSources/EventBasedAddOns
and referencedResources as part of snapshot
Collect HealthChecks as part of snapshot
Create CODE_OF_CONDUCT.md
Prepare for release v0.8.0
Prepare dev to main merge
Fix collection
Add CONTRIBUTING guidelines
Update CONTRIBUTING.md
Prepare dev to main merge
Fix for handling secrets
Rename show features to show addons
Kubernetes addons is the right terms to describe policies deployed by
sveltos in each managed cluster.
Advance pins
Advance pins
Fix README
Prepare for release v0.9.0
Prepare dev to main merge
Release v0.9.1
Prepare dev to main merge
Kustomize resources
Kustomize resources
Prepare for release v0.10.0
Prepare dev to main merge
Fix tag
Prepare for release v0.10.1
Prepare dev to main merge
Go version v1.20
Prepare for release v0.10.2
Prepare dev to main merge
Advance libsveltos pin
Prepare for release v0.11.0
Prepare dev to main merge
Rename AddonConstraint to AddonCompliance
Advance pins
- k8s v1.27.2
- controller-runtime v0.15.0
- cluster-api v1.5.0-beta.0
Prepare for release v0.12.0
Prepare dev to main merge
Add show resource command.
This command looks at all the HealthCheckReport instances and
display information about those.
For instance:
```
+-------------------------------------+--------------------------+----------------+-------------------------+----------------------------+
| CLUSTER | GVK | NAMESPACE | NAME | MESSAGE |
+-------------------------------------+--------------------------+----------------+-------------------------+----------------------------+
| default/sveltos-management-workload | apps/v1, Kind=Deployment | kube-system | calico-kube-controllers | All replicas 1 are healthy |
| | | kube-system | coredns | All replicas 2 are healthy |
| | | projectsveltos | sveltos-agent-manager | All replicas 1 are healthy |
+-------------------------------------+--------------------------+----------------+-------------------------+----------------------------+
```
Command as also option --full.
When set full resource will be displayed
```
Cluster: default/sveltos-management-workload
Object: object:
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
deployment.kubernetes.io/revision: "1"
creationTimestamp: "2023-07-07T11:23:41Z"
generation: 1
labels:
k8s-app: kube-dns
managedFields:
- apiVersion: apps/v1
fieldsType: FieldsV1
fieldsV1:
f:metadata:
f:labels:
.: {}
f:k8s-app: {}
f:spec:
...
```
Fix log level for two messages
Prepare for release v0.13.0
Prepare dev to main merge
Fix README
Prepare for release v0.14.0
Prepare dev to main merge
Post release v0.14.0
Prepare for release v0.15.0
Prepare dev to main merge
Post release v0.15.0
Merge dev to main
Post release v0.15.1
Post release v0.15.2
Add bug template
Allowing apiGroup lib.projectsveltos.io to update
Fix CI
Add check to make sure manifest.yaml is in sync
Prepare release v0.15.3
Prepare dev to main merge
Post release v0.15.3
Fix tag
Add command to register management cluster
Prepare for release v0.16.0
Prepare dev to main merge
Post release v0.16.0
Prepare for release v0.17.0
Prepare dev to main merge
Post release v0.17.0
Advance libsveltos and clusterapi pins
Prepare for release v0.18.0
Prepare dev to main merge
Post release v0.18.0
Pick long running job fix
issue described [here](projectsveltos/libsveltos@33de363)
Prepare dev to main merge
Post release v0.18.1
Update go modules
Update tag
Prepare for release v0.19.0
Prepare dev to main merge
Post release v0.19.0
Prepare for release v0.20.0
Prepare dev to main merge
Post release v0.20.0
Add support for Profiles
Commands updated:
- show addons
- show dryrun
- roolback
Also snapshot collects Profile instances as well.
Update tag to dev
Golang v1.21
Move golangci-lint to v1.55.2
Temporarily disable depguard
Prepare for release v0.21.0
Fix sveltosctl rbac
Prepare dev to main merge
Post release v0.21.0
Return an error if clusterSelector is not parsable
[bug](projectsveltos/libsveltos#260)
Bump libsveltos
Prepare for release v0.22.0
Prepare dev to main merge
Post release v0.22.0
Prepare for release v0.23.0
Prepare dev to main merge
Post release v0.23.0
show addons: fix cluster filter
When displaying deployed add-ons, command allows filtering
by namespace and cluster name.
Because of an issue, filtering by cluster name was returning no result.
This PR fixes it.
```
sveltosctl show addons
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
| CLUSTER | RESOURCE TYPE | NAMESPACE | NAME | VERSION | TIME | PROFILES |
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
| default/clusterapi-workload | helm chart | kyverno | kyverno-latest | 3.0.1 | 2024-02-14 16:12:47 +0100 CET | ClusterProfile/kyverno |
| default/clusterapi-workload | helm chart | nginx | nginx-latest | 0.14.0 | 2024-02-14 16:14:50 +0100 CET | ClusterProfile/kyverno |
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
```
Filter by namespace
```
sveltosctl show addons --namespace=default
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
| CLUSTER | RESOURCE TYPE | NAMESPACE | NAME | VERSION | TIME | PROFILES |
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
| default/clusterapi-workload | helm chart | kyverno | kyverno-latest | 3.0.1 | 2024-02-14 16:12:47 +0100 CET | ClusterProfile/kyverno |
| default/clusterapi-workload | helm chart | nginx | nginx-latest | 0.14.0 | 2024-02-14 16:14:50 +0100 CET | ClusterProfile/kyverno |
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
```
Filter by namespace and cluster name
```
sveltosctl show addons --namespace=default --cluster=clusterapi-workload
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
| CLUSTER | RESOURCE TYPE | NAMESPACE | NAME | VERSION | TIME | PROFILES |
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
| default/clusterapi-workload | helm chart | kyverno | kyverno-latest | 3.0.1 | 2024-02-14 16:12:47 +0100 CET | ClusterProfile/kyverno |
| default/clusterapi-workload | helm chart | nginx | nginx-latest | 0.14.0 | 2024-02-14 16:14:50 +0100 CET | ClusterProfile/kyverno |
+-----------------------------+---------------+-----------+----------------+---------+-------------------------------+------------------------+
```
Prepare for release v0.24.0
Prepare for release v0.25.0
Prepare dev to main merge
Post release v0.25.0
Prepare for release v0.26.0
Prepare dev to main merge
Post release v0.26.0
Prepare for release v0.27.0
Prepare dev to main merge
Post release v0.27.0
Introduce new command to generate kubeconfig
```
sveltosctl generate kubeconfig --help
Usage:
sveltosctl generate kubeconfig [options] [--namespace=<name>] [--serviceaccount=<name>] [--create] [--verbose]
--namespace=<name> The namespace of the ServiceAccount. If not specified, projectsveltos namespace will be used.
--serviceaccount=<name> The name of the ServiceAccount. If not specified, projectsveltos will be used.
--create If a ServiceAccount with enough permissions is already present, do not set this flag.
Sveltos will generate a Kubeconfig associated to that ServiceAccount.
If a ServiceAccount with cluster admin permissions needs to be created, use this option.
When this option is set, this command will create necessary resources:
1. namespace if not existing already
2. serviceAccount if not existing already
3. ClusterRole with cluster admin permission
4. ClusterRoleBinding granting the serviceAccount cluster admin permissions
5. TokenRequest for the ServiceAccount
Options:
-h --help Show this screen.
--verbose Verbose mode. Print each step.
Description:
The generate kubeconfig command will generate a Kubeconfig that can later on be used to register the cluster.
```
Enhance generate kubeconfig description
This PR also modifies the `sveltosctl register cluster` command
by allowing labels to be specified.
For instance
```
sveltosctl register cluster \
--namespace=monitoring \
--cluster=prod-cluster \
--kubeconfig=~/.kube/prod-cluster.config \
--labels=environment=production,tier=backend
```
This command registers a cluster named "prod-cluster" in
the "monitoring" namespace with labels "environment=production" and "tier=backend".
Register a cluster means creates a SveltosCluster instance in the "monitoring"
namespace named "prod-cluster" with labels "environment=production" and "tier=backend".
Allow user to pass expiration
This option allows you to specify the desired validity period
(in seconds) for the token requested when generating a kubeconfig
Allow to register a cluster with one command
The sveltosctl register command is used to register a cluster.
It requires a Kubeconfig to be passed.
If user has a kubeconfig with multiple contexts, the option __fleet-cluster-context__
allows to specify the context for the cluster to be managed.
So with default context pointing to the management cluster, following command will:
1. create a ServiceAccount in the managed cluster (using cluster-1 context)
2. grant this ServiceAccount cluster-admin permission
3. create a TokenRequest for such account and a Kubeconfig with bearer token from the TokenRequest
4. create a SveltosCluster in the management cluster (so using default context) and a Secret
with kubeconfig generated in the step above
```
sveltosctl register cluster --namespace=gcp --cluster=cluster-1 --fleet-cluster-context=cluster-1 --labels=k1=v1,k2=v2
```
Fix issue with register cluster
When fleet-cluster-context is used, after switching context, get the new kubeconfig
and use it to get host and CA data.
Always reset context
When register cluster is used with the fleet-cluster-context
always reset context to default. In case there is any failure
this will avoid changing the context without resetting it back
Prepare for release v0.28.0
Prepare dev to main merge
Post release v0.28.0
Prepare for release v0.29.0
Fix lint
Prepare dev to main merge
Post release v0.29.0
extended log command to set severity for components in managed clusters
handle log-level command for managed clusters
argument mismatch fix, undefined instance fix, correct GetClientForManagedCluster to GetClientForCluster
error handling in set function, logical check in update operations.
new approach
not enough arguments fix attempt
dealing with some errors
adding new params to test
creating GetDebuggingConfiguration and adding dynamic switching of clusterName
reverting changes in client.go as I misunderstood a part and maing change in debugging config file also removed default isntance
update params in test files
I thought there was a typo
unset missing the new params!
lint erros
new approach in utils.go
param order fix
reducing errors
removing extra params
removing nil return from collectlogconfig
trying a new fix
i dont think theres a point in returning the namespace and clustername in collectloglevelconfig
tried a new approach. should be better
making change sin debuggingconfig.go file
making the changes in the debbuging config test file
should fix the failing tests
mispell is on purpose, i keep forgetting!
fixing lint errors to make test vars constants
forgot to add change in utils.go debug config section
reverting change
adding clusterType and making additional changes
clusterproxy.GetSveltosKubernetesClient to get the client pointing to the managed cluster
might be an issue with k8sAccess redeclaration
Misunderstanding. This should be better
adding cluster type to req object
resolving some errors
adding logger as param
getting logger in a different way
logger is being tedious. debugging
fixing logger
new approach to get logger
getting logger new approach
new approach for logger
Passing it as a parameter
using correct GetKubernetesClient
getting logger
adding defaultInstanceName back
attempting to resolve error in test for debuggingconfig
attempt to fix test, different approach
reverting change
adding values instead of empty
testing fix
testing dc name
no need for addition in set_test
adding cluster type and namespace to DC
adding type and namespace to DC
Testing new approach for debuggingconfiguratgions
removing Type
new fixes
reveritng back to previous change.
reverting back to previous version
Changing expect test from Namespace to Name
Reading optional arguments correctly.
Syntax fix
Reading optional arguments correctly for show.go
Prepare for release v0.30.0
Prepare dev to main merge
Post release v0.30.0
Prepare for release v0.31.0
Prepare dev to main merge
Post release v0.31.0
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Defining a
Profile's orClusterProfile's selector with a wrong format, the Addon Controller crashes with the following stacktrace:This can be triggered as it follows
The error occurred on the last release (v0.21.0)
The text was updated successfully, but these errors were encountered: