Openshift 4 Blue green deployment at HomeLab

Try to deploy two openshift separate clusters in homelab. Remember in OCP3, 1 master and 2 workers can work perfectly in one single PC. Openshift 4 need more on resource and here is minimum requirement.

Minimum resource requirements

Each cluster machine must meet the following minimum requirements:

MachineOperating SystemvCPUVirtual RAMStorage

Bootstrap

FCOS

4

16 GB

120 GB

Control plane

FCOS

4

16 GB

120 GB

Compute

FCOS or RHEL 7.6

2

8 GB

120 GB    

https://docs.okd.io/latest/installing/installing_bare_metal/installing-bare-metal.html

For minimum installation require 1 Bootstrap, 3 control plane and 2 compute nodes. total vcpu: 20 and 80GB RAM. It is better to have 10 cores (1:2). 

Lab Environment:

Just upgrade one of my server to Intel(R) Xeon(R) CPU E5-2650L v3 @ 1.80GHz (2nd hand). It is good enough to run openshift in VMware. (12 cores, 96GB RAM)

URL: https://console-openshift-console.apps.green.hkopenshift.com/dashboards


On the other hand, I have two servers to create in RHV cluster with GlusterFS shared storage.

One PC is Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz with 32GB RAM about 10 years old. (4 cores)

Another supermicro Intel(R) Xeon(R) CPU D-1518 @ 2.20GHz CPU with 128GB RAM. Because limited CPU resource, it never finish installation on its own. (4 cores)

With total 8 cores, barely can install with some troubleshooting, fix timeout issue, CSR not auto signed, etc. For disk IO, the following tuning help a lot on disk performance. Finally, at least we have the console running and can deploy application on it.

Gluster volume: (applied using gluster volume set <volume> group virt): ◦disable quick-read, read-ahead, io-cache translators◦enable eager-lock, remote-dio volume parameters

https://access.redhat.com/sites/default/files/attachments/rhev-gluster-128-guests_0.pdf






Example of warning related to limited resource.
etcd cluster "etcd": 8.571428571428571 leader changes within the last 15 minutes. Frequent elections may be a sign of insufficient resources, high network latency, or disruptions by other components and should be investigated.

URL: https://console-openshift-console.apps.blue.hkopenshift.com/

"Blue" "green" deployment example. We can deploy application on two OCP and use the external load balancer (or DNS) to control traffic to the app. As the infrastructure(VMware and RHV) is total different and isolated, change on one platform will not affect another.


Small modify on example of hello world, just for demo.

https://github.com/bluesky255255/openshift-php-hello-world

#Openshift

#homelab

Deploy method IPI, Tested Version, OCP 4.6


Comments

Post a Comment

Popular posts from this blog

Openshift cert-manager integrate with ACME IdM on RHEL (Technical Preview)

Image
Explore ACME on IdM Redhat IPA (freeipa) AMCE in technical preview  The Automated Certificate Management Environment (ACME) service is now available in Identity Management (IdM) as a Technology Preview. <- RHEL 9.1 https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/9/html-single/9.1_release_notes/index#technology-preview_identity-management In RHEL 9.2. ACME now supports automatically removing expired certificates, but no mention about it is GA https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/9/html-single/9.2_release_notes/index#technology-preview_identity-management In RHEL 9.3 ACME available as a Technology Preview https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/9/html-single/9.3_release_notes/index#technology-preview_identity-management P.s. IdM is a software that comes with RHEL. only need one RHEL subscription. It supports replica (like MS domain controllers replication), but for simple setup, we may consider...
Read more

TimeSync using ansible and rhel-system-roles.timesync

Install   rhel-system-roles.timesync yum install rhel-system-roles -y After install  /usr/share/ansible/roles/rhel-system-roles.timesync [root@servera ~]# cat inventory   [all] server[a:e] [root@servera ~]# ansible -i inventory all -m ping   servere | SUCCESS => {     "changed": false,       "ping": "pong" } serverb | SUCCESS => {     "changed": false,       "ping": "pong" } servera | SUCCESS => {     "changed": false,       "ping": "pong" } serverd | SUCCESS => {     "changed": false,       "ping": "pong" } serverc | SUCCESS => {     "changed": false,       "ping": "pong" } [root@servera ~]# cat timesync.yml   --- - hosts: all   vars:     timesync_ntp_servers:       - hostname: {IP of NTP 1}     ...
Read more

Remove HPE iLO license

Image
It spend me sometime to reserch how to remove iLO advance license and use HPE OneView Advanced. Here is the solution. 1. download HP Lights-Out configuration Utility https://support.hpe.com/hpsc/swd/public/detail?sp4ts.oid=null&swItemId=MTX_8abe539b67bf46978e8f84acb8&swEnvOid=4184#tab1 2. run the command with xml "C:\Program Files (x86)\Hewlett Packard Enterprise\HP Lights-Out Configuration Utility\HPQLOCFG.exe" -s 10.10.100.43 -f C:\Users\chengwi\Desktop\ilo.xml HP Lights-Out Configuration Utility - HPQLOCFG v5.2.0.0 dated 12/18/2017 (c) 2013, 2017 Hewlett Packard Enterprise Development LP Detecting iLO.... Negotiated cipher:168-bit TripleDes with 160-bit Sha1 and 1024-bit RsaKeyX .... Script succeeded for IP:10.10.100.43:443 ilo.xml      
Read more