Dataset Viewer
input
stringclasses 6
values | output
stringclasses 6
values |
|---|---|
K3s cluster bootstrap
Bootstrap with Puppet
Puppet configuration
Cluster initialization
CI job
Input
Outputs
Operations
Create DNS record for cluster access
Create config files for k3s
Start/enable k3s
Output display
Bootstrap with Puppet
Puppet configuration
Inlcude the kbi::k3s module, and set the following parameters:
Cluster initialization
Run puppet (or let it run) on all nodes.
With reference to the above context, What does the CI job do?
|
Create DNS record for cluster access
|
You can either:
do that in order: by beginning by the initiator controller, then running puppet on all other controllers, then all workers once it has finished.
Q: What is the first step to do?
Referring to the passage above, the correct answer to the given question is
|
beginning by the initiator controller
|
I know that the answer to the question "What is the main point of the text?" is in "not caring and running it on all of them at the same time: some services will fail at first but will retry until the initiator has finished, ultimately all nodes will join the cluster.". Can you tell me what it is?
|
not caring and running it on all of them at the same time
|
CI job
Input
customer : String
Ex: ig1
project : String
Ex: kbilab
server_nodes : Array[String]
A list of FQDNs of controller nodes
Ex:
agent_nodes : Array[String]
A list of FQDNs of worker nodes
Ex:
k3s_version : String
Ex: 1.28.3+k3s2-ig1~1+bookworm
cluster_cidr : Optional[String]
Default: 10.96.0.0/16
service_cidr : Optional[String]
Default: 10.97.0.0/16
local_storage_path : Optional[String]
Default: /data/k3s/local-storage
1
kbi::k3s::version: "1.28.3+k3s2-ig1~1+bookworm"
2
kbi::k3s::token: "************" #generate a random password
3
kbi::k3s::registration_record: "ig1-kbilab-controllers.vtr.sadm.ig-1.net" #use the common registration address previously created
4
kbi::k3s::mode: server #on controllers
5
#kbi::k3s::mode: agent #on workers
6
#kbi::k3s::initiator: true #on one controller only (typically the first)
1
[ "ig1-kbilab-controller-01.vtr.sadm.ig-1.net",
2
"ig1-kbilab-controller-02.vtr.sadm.ig-1.net",
3
"ig1-kbilab-controller-03.vtr.sadm.ig-1.net"]
1
[ "ig1-kbilab-worker-01.vtr.sadm.ig-1.net",
2
"ig1-kbilab-worker-02.vtr.sadm.ig-1.net"]
Initiator name and token will be automatically generated and handled by our CI.
Q: What is the default local storage path?
A:
|
/data/k3s/local-storage
|
Refer to the passage below and answer the following question:
Passage: Special cases to handle:
Outputs
Operations
Create DNS record for cluster access
Push a DNS record to Route53 with an RR to all controllers.
Question: What is one special case to handle?
|
Outputs
|
Create config files for k3s
for each server:
/etc/rancher/k3s/config.yaml :
/etc/default/k3s :
for each agent:
/etc/rancher/k3s/config.yaml :
/etc/default/k3s :
Start/enable k3s
enable service k3s.service
start service k3s.service
Output display
DNS record of all controllers
聽
聽
only 1 controller
1
---
2
cluster-cidr: <cluster_cidr>
3
service-cidr: <service_cidr>
4
disable:
5
- traefik
6
- servicelb
7
- metrics-server
8
default-local-storage-path: <local_storage_path>
9
flannel-ipv6-masq: true
10
tls-san:
11
- <server_hostname>
12
- <project>.<customer>.kbi.ig-1.net
13
kube-apiserver-arg:
14
- enable-admission-plugins=PodTolerationRestriction
1
K3S_MODE=server
2
K3S_TOKEN=<token_randomly_generated>
3
K3S_ADDITIONAL_FLAGS='--cluster-init' # if first controller
4
K3S_ADDITIONAL_FLAGS='--server <server_nodes[0]>' # else
1
---
1
K3S_MODE=agent
2
K3S_TOKEN=<token_randomly_generated>
3
K3S_ADDITIONAL_FLAGS='--server <server_nodes[0]>'
Q: What is the name of the file that should be created for each server?
Referring to the passage above, the correct answer to the given question is
|
/etc/rancher/k3s/config.yaml
|
README.md exists but content is empty.
- Downloads last month
- 3