Nicolas Vazquez 6adfda2818

CKS Enhancements (#9102) ...

CKS Enhancements:

* Ability to specify different compute or service offerings for different types of CKS cluster nodes – worker, master or etcd

* Ability to use CKS ready custom templates for CKS cluster nodes

* Add and Remove external nodes to and from a kubernetes cluster

Co-authored-by: nvazquez <nicovazquez90@gmail.com>

* Update remove node timeout global setting

* CKS/NSX : Missing variables in worker nodes

* CKS: Fix ISO attach logic

* CKS: Fix ISO attach logic

* address comment

* Fix Port - Node mapping when cluster is scaled in the presence of external node(s)

* CKS: Externalize control and worker node setup wait time and installation attempts

* Fix logger

* Add missing headers and fix end of line on files

* CKS Mark Nodes for Manual Upgrade and Filter Nodes to add to CKS cluster from the same network

* Add support to deploy CKS cluster nodes on hosts dedicated to a domain

---------

Co-authored-by: Pearl Dsilva <pearl1594@gmail.com>

* Support unstacked ETCD

---------

Co-authored-by: nvazquez <nicovazquez90@gmail.com>

* Fix CKS cluster scaling and minor UI improvement

* Reuse k8s cluster public IP for etcd nodes and rename etcd nodes

* Fix DNS resolver issue

* Update UDP active monitor to ICMP

* Add hypervisor type to CKS cluster creation to fix CKS cluster creation when External hosts added

* Fix build

* Fix logger

* Modify hypervisor param description in the create CKS cluster API

* CKS delete fails when external nodes are present

* CKS delete fails when external nodes are present

* address comment

* Improve network rules cleanup on failure adding external nodes to CKS cluster

* UI: Fix etcd template was not honoured

* UI: Fix etcd template was not honoured

* Refactor

* CKS: Exclude etcd nodes when calculating port numbers

* Fix network cleanup in case of CKS cluster failure

* Externalize retries and inverval for NSX segment deletion

* Fix CKS scaling when external node(s) present in the cluster

* CKS: Fix port numbers displayed against ETCD nodes

* Add node version details to every node of k8s cluster - as we now support manual upgrade

* Add node version details to every node of k8s cluster - as we now support manual upgrade

* update column name

* CKS: Exclude etcd nodes when calculating port numbers

* update param name

* update param

* UI: Fix CKS cluster creation templates listing for non admins

* CKS: Prevent etcd node start port number to coincide with k8s cluster start port numbers

* CKS: Set default kubernetes cluster node version to the kubernetes cluster version on upgrade

* CKS: Set default kubernetes cluster node version to the kubernetes cluster version on upgrade

* consolidate query

* Fix upgrade logic

---------

Co-authored-by: nvazquez <nicovazquez90@gmail.com>

* Fix CKS cluster version upgrade

* CKS: Fix etcd port numbers being skipped

* Fix CKS cluster with etcd nodes on VPC

* Move schema and upgrade for 4.20

* Fix logger

* Fix after rebasing

* Add support for using different CNI plugins with CKS

* Add support for using different CNI plugins with CKS

* remove unused import

* Add UI support and list cni config API

* necessary UI changes

* add license

* changes to support external cni

* UI changes

* Fix NPE on restarting VPC with additional public IPs

* fix merge conflict

* add asnumber to create k8s svc layer

* support cni framework to use as-numbers

* update code

* condition to ignore undefined jinja template variables

* CKS: Do not pass AS number when network ID is passed

* Fix deletion of Userdata / CNI Configuration in projects

* CKS: Add CNI configuration details to the response and UI

* Explicit events for registering cni configuration

* Add Delete cni configuration API

* Fix CKS deployment when using VPC tiers with custom ACLs

* Fix DNS list on VR

* CKS: Use Network offering of the network passed during CKS cluster creation to get the AS number

* CKS cluster with guest IP

* Fix: Use control node guest IP as join IP for external nodes addition

* Fix DNS resolver issue

* Improve etcd indexing - start from 1

* CKS: Add external node to a CKS cluster deployed with etcd node(s) successfully

* CKS: Add external node to a CKS cluster deployed with etcd node(s) successfully

* simplify logic

* Tweak setup-kube-system script for baremetal external nodes

* Consider cordoned nodes while getting ready nodes

* Fix CKS cluster scale calculations

* Set token TTL to 0 (no expire) for external etcd

* Fix missing quotes

* Fix build

* Revert PR 9133

* Add calico commands for ens35 interface

* Address review comments: plan CKS cluster deployment based on the node type

* Add qemu-guest-agent dependency for kvm based templates

* Add marvin test for CKS clusters with different offerings per node type

* Remove test tag

* Add marvin test and fix update template for cks and since annotations

* Fix marvin test for adding and removing external nodes

* Fix since version on API params

* Address review comments

* Fix unit test

* Address review comments

* UI: Make CKS public templates visible to non-admins on CKS cluster creation

* Fix linter

* Fix merge error

* Fix positional parameters on the create kubernetes ISO script and make the ETCD version optional

* fix etcd port displayed

* Further improvements to CKS  (#118)

* Multiple nics support on Ubuntu template

* Multiple nics support on Ubuntu template

* supports allocating IP to the nic when VM is added to another network - no delay

* Add option to select DNS or VR IP as resolver on VPC creation

* Add API param and UI to select option

* Add column on vpc and pass the value on the databags for CsDhcp.py to fix accordingly

* Externalize the CKS Configuration, so that end users can tweak the configuration before deploying the cluster

* Add new directory to c8 packaging for CKS config

* Remove k8s configuration from resources and make it configurable

* Revert "Remove k8s configuration from resources and make it configurable"

This reverts commit d5997033ebe4ba559e6478a64578b894f8e7d3db.

* copy conf to mgmt server and consume them from there

* Remove node from cluster

* Add missing /opt/bin directory requrired by external nodes

* Login to a specific Project view

* add indents

* Fix CKS HA clusters

* Fix build

---------

Co-authored-by: Nicolas Vazquez <nicovazquez90@gmail.com>

* Add missing headers

* Fix linter

* Address more review comments

* Fix unit test

* Fix scaling case for the same offering

* Revert "Login to a specific Project view"

This reverts commit 95e37563f48573780b07a038a7f48c0bc04e9b64.

* Revert "Fix CKS HA clusters" (#120)

This reverts commit 8dac16aa359faa6500ea1e1ce548169cfd08331a.

* Apply suggestions from code review about user data

Co-authored-by: Suresh Kumar Anaparti <sureshkumar.anaparti@gmail.com>

* Update api/src/main/java/org/apache/cloudstack/api/command/user/userdata/BaseRegisterUserDataCmd.java

Co-authored-by: Suresh Kumar Anaparti <sureshkumar.anaparti@gmail.com>

* Refactor column names and schema path

* Fix scaling for non existing previous offering per node type

* Update node offering entry if there was an existing offering but a global service offering has been provided on scale

---------

Co-authored-by: Pearl Dsilva <pearl1594@gmail.com>
Co-authored-by: Daan Hoogland <daan@onecht.net>
Co-authored-by: Suresh Kumar Anaparti <sureshkumar.anaparti@gmail.com>

2025-06-19 11:00:42 +05:30

..

22.04

CKS Enhancements (#9102)

2025-06-19 11:00:42 +05:30

build.sh

CKS Enhancements (#9102)

2025-06-19 11:00:42 +05:30