Hi Bernd, Thankyou so much for your time for guidance, highly appreciated!
As far as deployment is concerned, I started deploying openstack-Zed release using ansible in Docker containers:
https://docs.openstack.org/project-deploy-guide/kolla-ansible/zed/.
Initially, I setup multinode cluster 1x controller, 1x compute, 1x storage and followed the guide. After adding dependencies packages of python, kolla and ansible, I created user with same nodes for ssh connectivity. The ansible was able to ping all the nodes through controller node by:
Ansible -I multinode all -m ping.
For services to run properly, I used firewalld and created rules to allow traffic from different ports as below:
sudo systemctl enable firewalld sudo firewall-cmd --zone=public --add-port=5000/8774/96960/tcp --permanent
I have attached multinode inventory and ansible connectivity for your verification.
For the placement services, I can work on it. I would appreciate if you guide and help me to setup the desired multinode cluster and share some learning architectural concepts with me. Looking forward to a pleasant cooperation!
Thanks, -----Original Message----- From: berndbausch@gmail.com berndbausch@gmail.com Sent: Tuesday, November 21, 2023 5:52 AM To: 'Asma Naz Shariq' asma.naz@techavenue.biz; openstack-discuss@lists.openstack.org Subject: RE: openstack-discuss - Openstack Kolla-ansible Nova Compute - Error | Urgent
You need to find out whether there is connectivity from the compute node to the services running on the controller. My guess is that there is no connectivity, and you will then have to find out why.
This includes connectivity to RabbitMQ running on the controller. In addition, something seems to be amiss with the Placement service as well.
Services may be unreachable for a number of reasons. On the compute node, I would check for things like incorrect name resolution, incorrect IP addresses, incorrect port numbers. On the controller: On which IP addresses and port numbers are services listening? And cloud-wide: Do the endpoints registered with Keystone correspond to the actual IP addresses, port numbers and/or URLs?
Tools to check connectivity: ping, netcat (also known as ncat or nc), curl.
There is a WARNING in the log you are sharing - "Failure to contact the endpoint at http://10.10.10.30:8780 for discovery". It's only a warning but still - it may provide a clue to the problem. Does this IP address exist? Is anything listening at that port?
In my case, learning OpenStack required learning about networking, which I was quite ignorant about at that time (I still am, but less so). I am sure that the majority of problems in OpenStack clouds are caused by some incorrect network configuration.
By the way, how did you deploy this cloud? Do you have any suspicious messages in the deployment log files?
Bernd.
-----Original Message----- From: Asma Naz Shariq asma.naz@techavenue.biz Sent: Monday, November 20, 2023 8:19 PM To: berndbausch@gmail.com; openstack-discuss@lists.openstack.org Subject: RE: openstack-discuss - Openstack Kolla-ansible Nova Compute - Error | Urgent
Hi Berndbausch,
Thankyou for your reply. I tried to resolve the compute node host in /etc/hosts file through (ip : computenode-virtual-machine). I can check that logs have been changed for nova-compute log on compute node as attached. I am unable to find the solution. Can you check this for me and guide me how to resolve these issues. I would appreciate if you schedule a session for it. T -----Original Message----- From: berndbausch@gmail.com berndbausch@gmail.com Sent: Monday, November 20, 2023 2:24 PM To: 'Asma Naz Shariq' asma.naz@techavenue.biz; openstack-discuss@lists.openstack.org Subject: RE: openstack-discuss -Nova Compute - Error | Urgent
The compute node tries to register with Nova on the controller but can't communicate with it. The problem appears to be at RabbitMQ, which says it can't reach the controller.
First things you should check: Unsynchronized clocks and incorrect configuration of name resolution on the compute node.
-----Original Message----- From: Asma Naz Shariq asma.naz@techavenue.biz Sent: Monday, November 20, 2023 3:50 PM To: openstack-discuss@lists.openstack.org Subject: RE: openstack-discuss -Nova Compute - Error | Urgent
Hi Openstack Community,
I am deploying Openstack zed release Multinode cluster with 1x Controller and 1x compute and 1x storage. At the time of deployment, I encountered a below error for failed to register the compute service:
TASK [nova-cell: Fail if nova-compute service failed to register] *********************************** fatal: [X.X.X.X]: FAILED! => {"changed": false, "msg": "The Nova compute service failed to register itself on the following hosts: computenode-virtual-machine"} fatal: [X.X.X.X]: FAILED! => {"changed": false, "msg": "The Nova compute service failed to register itself on the following hosts: computenode-virtual-machine"}
The logs in nova-compute.log at compute node state the ERROR: 2023-11-20 11:54:32.008 7 ERROR oslo. messaging._drivers.impl_rabbit [None req-f8176572-75bf-4c27-a9be-e096cb43e2e3 - - - - - -] Connection failed: [Errno 113] EHOSTUNREACH (retrying in 31.0 seconds): OSError: [Errno 113] EHOSTUNREACH.
Can anyone guide me how to resolve this issue?
Thanks
-----Original Message----- From: openstack-discuss-request@lists.openstack.org openstack-discuss-request@lists.openstack.org Sent: Monday, November 20, 2023 10:16 AM To: openstack-discuss@lists.openstack.org Subject: openstack-discuss Digest, Vol 61, Issue 70
Send openstack-discuss mailing list submissions to openstack-discuss@lists.openstack.org
To subscribe or unsubscribe via email, send a message with subject or body 'help' to openstack-discuss-request@lists.openstack.org
You can reach the person managing the list at openstack-discuss-owner@lists.openstack.org
When replying, please edit your Subject line so it is more specific than "Re: Contents of openstack-discuss digest..."
Today's Topics:
1. Re: about magnum capi for production (Nguyễn Hữu Khôi)
----------------------------------------------------------------------
Message: 1 Date: Mon, 20 Nov 2023 12:16:11 +0700 From: Nguyễn Hữu Khôi nguyenhuukhoinw@gmail.com Subject: Re: about magnum capi for production To: Satish Patel satish.txt@gmail.com Cc: OpenStack Discuss openstack-discuss@lists.openstack.org Message-ID: CABAODRfcEA38QRLCpr_DtC5Dbxmr2x=G9nD8hENptAdh0Z0QXg@mail.gmail.com Content-Type: multipart/alternative; boundary="000000000000068859060a8e96ec"
Hello guys.
This is a doc which is about how to make Magnum CAPI to work with Openstack deployed by Kolla Ansible..
https://github.com/ngyenhuukhoi/openstack-note/blob/main/Magnum%20Cluster%20...
Feel free to ask. If I can help, I will.
Nguyen Huu Khoi
On Mon, Nov 20, 2023 at 12:09 PM Nguyễn Hữu Khôi nguyenhuukhoinw@gmail.com wrote:
Hello. It uses prebuild Ubuntu to provision k8s cluster,
On Mon, Nov 20, 2023 at 11:47 AM Satish Patel satish.txt@gmail.com wrote:
Hi Folks,
Tell me if I am wrong. CAPI is just kubernetes images to build k8s clusters instead using coreOS and install a bunch of stuff via heat. In documentation all I can see are imported images. Reference doc: https://vexxhost.github.io/magnum-cluster-api/user/getting-started/
Do I need to install anything else to make it work?
On Fri, Nov 17, 2023 at 11:30 AM Mohammed Naser mnaser@vexxhost.com wrote:
Thanks! Happy to see that it is working smoothly 😊
Appreciate any docs, comments or issues.
*From: *Nguyễn Hữu Khôi nguyenhuukhoinw@gmail.com *Date: *Friday, November 17, 2023 at 8:57 AM *To: *Mohammed Naser mnaser@vexxhost.com *Cc: *OpenStack Discuss openstack-discuss@lists.openstack.org *Subject: *Re: about magnum capi for production
Hello guys.
I have successfully implemented Magnum CAPI on my openstack system. My deployment tool for Openstack is Kolla-Ansible. It worked greatly, I am testing with create, delete and resize k8s cluster. It is smooth when compared with Heat Driver. I will make some docs for new users to use it.
Thank you vexxhost.com and contributors for very nice production.
I will keep testing and reporting to help it be much better in the future.
Nguyen Huu Khoi
On Tue, Nov 14, 2023 at 5:50 AM Nguyễn Hữu Khôi < nguyenhuukhoinw@gmail.com> wrote:
Hello.
I get it now;
Thank you very much.
Nguyen Huu Khoi
On Mon, Nov 13, 2023 at 11:20 PM Mohammed Naser mnaser@vexxhost.com wrote:
Hi there,
Nope – it is simply a different driver, so if you try to create Fedora CoreOS clusters – it’ll use the old heat driver. Otherwise, it will use the Cluster API if the image is marked as Ubuntu
Thanks!
Mohammed
*From: *Nguyễn Hữu Khôi nguyenhuukhoinw@gmail.com *Date: *Monday, November 13, 2023 at 1:58 AM *To: *Mohammed Naser mnaser@vexxhost.com *Cc: *OpenStack Discuss openstack-discuss@lists.openstack.org *Subject: *Re: about magnum capi for production
Hello Mohammed,
I found that
https://vexxhost.github.io/magnum-cluster-api/user/getting-started/# creating
I would like to know that if we use cluster api, does it overwrite heat when deploying new k8s cluster from horizon?
Nguyen Huu Khoi
On Mon, Oct 30, 2023 at 7:49 PM Nguyễn Hữu Khôi < nguyenhuukhoinw@gmail.com> wrote:
Awesome, thank you so much!
Nguyen Huu Khoi
On Mon, Oct 30, 2023 at 7:31 PM Mohammed Naser mnaser@vexxhost.com wrote:
Sounds good, thank you.
We also have a channel on the Kubernetes Slack that is relatively active if you need help — it’s in the README :)
Thanks!
Mohammed
Get Outlook for iOS https://aka.ms/o0ukef
*From:* Nguyễn Hữu Khôi nguyenhuukhoinw@gmail.com *Sent:* Monday, October 30, 2023 8:23:07 AM *To:* Mohammed Naser mnaser@vexxhost.com *Cc:* OpenStack Discuss openstack-discuss@lists.openstack.org *Subject:* Re: about magnum capi for production
Hello.
Thank you very much. I am planning to use Magnum CAPI for my system. I will feedback later.
Nguyen Huu Khoi
On Mon, Oct 30, 2023 at 7:03 PM Mohammed Naser mnaser@vexxhost.com wrote:
Hi there!
It works out of the box since the goal is to have an API that is compatible so that all the ecosystem tools such as Terraform, Ansible, CLI, Horizon all continue to work just fine!
We’ve even made some quality of life improvements to the Magnum plugin like letting you download KUBECONFIG from it.
Thanks
Mohammed
Get Outlook for iOS https://aka.ms/o0ukef
*From:* Nguyễn Hữu Khôi nguyenhuukhoinw@gmail.com *Sent:* Monday, October 30, 2023 1:09:10 AM *To:* Mohammed Naser mnaser@vexxhost.com *Cc:* OpenStack Discuss openstack-discuss@lists.openstack.org *Subject:* Re: about magnum capi for production
Hello,
Thank you for your sharing, Could you tell me how we use it with Horizon or we need custom Horizon to use Magnum CAPI.
Nguyen Huu Khoi
On Mon, Oct 30, 2023 at 11:10 AM Mohammed Naser mnaser@vexxhost.com wrote:
Hi there.
We deploy this driver in production in our public cloud as well as many private clouds and as part of Atmosphere and we’ve been fully maintaining it on our side here.
Let me know if you have any questions
Thanks
Mohammed
Get Outlook for iOS https://aka.ms/o0ukef
*From:* Nguyễn Hữu Khôi nguyenhuukhoinw@gmail.com *Sent:* Sunday, October 29, 2023 10:22:22 PM *To:* OpenStack Discuss openstack-discuss@lists.openstack.org *Subject:* about magnum capi for production
Hello guys.
Could you share with me some experience about using Magnum CAPI in production?
https://github.com/vexxhost/magnum-cluster-api
Thank you so much.
Nguyen Huu Khoi
