Common Q&A document.
Here are some common install/run questions.
status meaning:
:construction: under develop :memo: in disscussion :raising_hand: help wanted :white_check_mark: resolved
Root permission rela
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
root User |
sudo sealos run not works well |
All command must run at root user, currently donot support run as sudoers or rootless |
#1386, #690 | :construction: |
Fix:point_right:: using root user to exec commands.
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
| Single Cluster | sealos run without any ips |
Currently sealos run must spec at least one master and one node |
#1353 #1064 | #1396 | :white_check_mark: |
Fix:point_right:: using atleast two machine(or multipass to start vms quickly)
System environment rela
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
invalid argument "never" for " --pull" flag: strconv. ParseBool: parsing "never": invalid syntax |
Cannot run sealos run command |
System has buildah installed but version do not match sealos's requirement |
#1961 | :white_check_mark: |
Fix:point_right:: Delete System's buildah and using sealos's build-in version.
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
| Cannot install sealos cluster when offline | Exec apparmor_parser file not found |
Default Set disableApparmor=false, Update Image version |
#1755 | labring/cluster-image@8b4bb02, labring/cluster-image#211 | :white_check_mark: |
Re-Run sealos fail on a sealos reseted cluster |
Re-Run failed with container runtime is not running or token exit status 1 and similar errors |
sealos reset donot clearly cleaned all System enviroments |
#1866, #1998 | #1936 | :white_check_mark: |
Fix:point_right:: Upgrade to 4.1.4-rc1 and above!
Network connective rela
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
| Validate Fail | couldn't validate the identity of the API Server, something like 10.103.97.2:6443 fail |
due to connective block between masters and nodes | :memo: | ||
Offline run *.tar.gz |
Install from *.tar.gz offline image package,make offline installation easier |
#1770 | #1861 | :white_check_mark: | |
Support customize hostname under cluster file |
It is useful to specific hostname under cluster file when there are many nodes. | #1729, #1449 | :raising_hand: | ||
Support extenal etcd cluster |
Use kubeadm to customize |
#1612 | :raising_hand: |
Fix:point_right:: double check masters and nodes can access each other at any port!
Not only ssh(22) port but also api-server(6443) and many others as different serivces/pods needs different ports open. Especially checking security group when using aws/aliyun infras.
Currently we suggest allow all2all under internal/private vpc/subnets.
Storage requirements rela
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
| Download Image Fail | sealos run CLusterImage fail |
lack of disk spaces | :raising_hand: | ||
| Download Image Fail | sealos build still fail while pulling an image from the private registry |
lack of memory due to directio bug | #1314 | :construction: | |
| Change install directory | sealos default install on root home dir but want to change this location |
small root directory size | #1592 | :white_check_mark: |
Fix:point_right:: add more storages and/or memories!
After install and run
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
| How to change masters/nodes ip | there is no document or command to support change nodes ips | #1406 | :raising_hand: | ||
| How to access private registry | you can access private registry directly | #1412 | #1426 | 🚧 | |
--cmd not work |
--cmd except to merge build-in commands |
Upgrade to 4.1.4-rc1 and above | #1969 | :white_check_mark: | |
error converting YAML to JSON |
apply clusterfile shows error | Rename *.yaml.tmpl file to *.yaml.tmpl.bak, as .tmpl suffix file will be rendered by default |
#1454 | :white_check_mark: | |
| Manage masters/nodes fail at running clusters | Mount data is missing, need update | #1402 | #1429 | :construction: | |
| Some masters/nodes fail to start after reboot machine | Hosts data is missing, need update | #1619, #1621, #1894 | :construction: | ||
Running sealos other than master0 |
many codes need to be update | #1276,#1567 | #1292 | :white_check_mark: |
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
Add new masters fail |
Time created by sealctl token do not synced or expired |
#1783 | #1936 | :white_check_mark: |
Fix:point_right:: Upgrade to 4.1.4-rc1 and above!
Cluster image rela
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
Build image with digest failed |
Some images like cilium, have digest inside, sealos build will fail |
It's image-shim witch needs to deal with this problem | #1460 | #1466 | :white_check_mark: |
| Pull arm image at x86 platform | #1929 | :white_check_mark: | |||
| Pull image do not update local's | Default pull policy is hardcoded missing, witch do not update for local's already pulled image |
#1877 | #1921 | :white_check_mark: |
Fix:point_right:: Upgrade to 4.1.4-rc1 and above!
| Summary | Description | Cause | Issue | PR | Status |
|---|---|---|---|---|---|
| Simplest way to set custom parameters | When one want to add custom kubeadm init parameters, must create an Clusterfile, and put it into Clusterfile | Add Clusterfile into image to run as default Clusterfile | #1459 | #1591 | :white_check_mark: |
| How to use sealos build-in private registry | docs/design/private-registry | #1412 | :white_check_mark: | ||
ClusterImage ingress-nginx pod keeps pending |
clusterfile nodeselector error | #1405 | :white_check_mark: | ||
Merge ClusterImage using env |
One can set env vars pass to ClusterImage to override tepmlates |
#1724 | #1480,#1733 | :white_check_mark: | |
kuboard install fail |
Use kuboard with tested kubernetes version |
#1909,eip-work/kuboard-press#328 | :white_check_mark: | ||
No matches for kind PodSecurityPolicy |
Install kubernetes:v1.25 with calico |
When install k8s v1.25, using labring/calico:v3.24.1 |
#1775 | :white_check_mark: | |
| sealos can't load docker image | It's sealos's design, witch deal with ClusterImage: build-example-cloudimage/中文 | #1655 | :white_check_mark: |
Before contributing to docs, please check out docs contributing guide.
Before contributing to docs, please check out docs contributing guide.
https://github.com/labring/sealos/issues/2008#issuecomment-1305269715
https://github.com/labring/sealos/issues/2757
https://github.com/labring/sealos/issues/2370
Hello, may I ask what kind of work is required for this problem?