Re: Slow Metrics output in GUI

[Joan g]

just cleaning up the vm_stats table in cloudstack 'cloud' db.

> truncate table vm_stats;

and setting  vm.stats.max.retention.time to a lower value addressed our
issues.

Joan

On Tue, Feb 27, 2024 at 12:19 AM Andrei Mikhailovsky
 wrote:

> Interesting.
>
> Joan, do you mind sharing how you are doing it?
>
> Thanks
>
> - Original Message -
> > From: "Joan g" 
> > To: "users" 
> > Sent: Monday, 26 February, 2024 18:06:58
> > Subject: Re: Slow Metrics output in GUI
>
> > I am facing the same problem in my 4.17.2 version. We are manually
> clearing
> > the stats table to  make the instance list page load faster :(
> >
> >
> > Joan
> >
> > On Mon, 26 Feb, 2024, 22:24 Andrei Mikhailovsky,
> 
> > wrote:
> >
> >> Hello everyone,
> >>
> >> My setup: ACS 4.18.1.0 on Ubuntu 20.04.6. Two management servers and
> mysql
> >> active-active replication.
> >>
> >>
> >> I seem to have a very slow response on viewing vms. It takes about 20
> >> seconds for the vm data to show when I click on any vm under Compute >
> >> Instances. When I click on various vm tabs (like NICs, Disks, Details,
> etc)
> >> the only tab that takes about 15-20 seconds to refresh is the Metrics
> tab.
> >> When the spinner stops I get the following message: "No data to show for
> >> the selected period." Also this information is shown in red colour: The
> >> Control Plane Status of this instance is Offline. Some actions on this
> >> instance will fail, if so please wait a while and retry. When I click on
> >> the 12 or 24 hours tab it takes a bit of time, but it does show the
> tables
> >> and the message in red colour is not shown.
> >> On mysql server I see the mysql process is using over 100% cpu (with 0%
> >> iowait) while ACS tries to retrieve the Metrics data. Also, the
> >> cloudstack-management server cpu usage goes to 200-400%.
> >>
> >>
> >> I've tried all the obvious (restarting management servers, stopping one
> of
> >> the management servers, restarting host servers).
> >>
> >> Does anyone know what is the issue? why does it take so long to retrieve
> >> the vm data and metrics? I don't remember having this problem before
> 4.18.
> >>
> >> Many thanks for any pointers.
> >>
> >> cheers
> >>
> >> Andrei
> >>
> >>
> >>
> >>
> >>
> >>
> >>
>

Re: GPU discovery in the hypervisor

[Nux]

This sounds foreign to me, afaik GPU support is limited to certain (old) 
NVIDIA Grid cards on Xenserver Enterprise.
Modern GPUs are not supported out of the box, although of course many 
here do use them by means of custom xml/groovy scripts.


How you detect them, no idea, let's see how other users do it, if they 
care to share.


On 2024-02-26 18:00, Douglas Oliveira wrote:

Hello,
How does the GPU discovery process work on the hypervisor with SC,
something similar to what Opennebula does? (through lspci)
I currently have a service offering created via API for an Nvidia A16 
GPU,

which does not work because it is informed that there are no hosts
available to serve the resource. So I'm unsure whether what doesn't 
work is

the service offering or the non-detection of the GPU on the host.

Regards

Re: new committer: Vishesh Jindal (vishesh)

[Nux]

Congrats & well deserved!

On 2024-02-26 14:05, Daan Hoogland wrote:

users and devs,

The Project Management Committee (PMC) for Apache CloudStack
has invited Vishesh Jindal to become a committer and we are pleased
to announce that they have accepted.

Being a committer enables easier contribution to the
project since there is no need to go via the patch
submission process. This should enable better productivity.

Please join me in congratulating Vishesh.

Re: Cloudstack with Managed Databases?

[Hunter Yap]

Thanks for all the feedback.

Actually what i was kind of looking at achieving is something similar to
our own 'Managed Database' like AWS Aurora RDMS where the OS upgrades,
database upgrades, high availability, replication etc is all handled by the
cloud provider (Which we are also trying to figure out how to achieve). I
think these use kubernetes of some sort? Im not sure.

Regards,
Hunter

On Mon, 26 Feb 2024 at 20:28, Kiran Chavala 
wrote:

> +1 for using Cloud-init to  install and configure your DBMS.
>
> Or you can use packer plugin for cloudstack to create golden image
> template with DBMS
>
>
> https://developer.hashicorp.com/packer/integrations/hashicorp/cloudstack/latest/components/builder/cloudstack
> https://copyprogramming.com/howto/how-to-setup-mysql-with-cloud-init
>
>
> Regards
> Kiran
>
> From: Jayanth Reddy 
> Date: Monday, 26 February 2024 at 8:15 AM
> To: users@cloudstack.apache.org 
> Subject: Re: Cloudstack with Managed Databases?
> Hello,
> One of the ways I think of is to make use of the cloud-init's
> functionalities to install and configure your DBMS. However you may have
> less view of the DBMS later which might not exactly fit into "managed".
>
> I've seen people have their own CMP and handling all the integration there.
>
> Thanks,
> Jayanth
>
>
>
>
> 
> From: Hunter Yap <123qwqqw...@gmail.com>
> Sent: Monday, February 26, 2024 8:04:39 am
> To: users@cloudstack.apache.org 
> Subject: Cloudstack with Managed Databases?
>
> Hi Guys,
>
> We are exploring offering Managed Databases as a service on our Cloudstack
> Public Cloud.
>
> Has anyone done this before? What method did you use and what was the
> experience like?
>
> Regards,
> Hunter
>

Advice Needed: Connecting VPCs Across Different Account

[Palash Biswas]

Hi everyone,

I'm seeking advice on the best method to connect VPCs across different
accounts. Should I use a private gateway or a site-to-site VPN?

I have two accounts and want to establish communication between the VPC in
Account A and the VPC in Account B.

Thanks,
Palash Biswas

Re: Advice Needed: Connecting VPCs Across Different Account

[Simon Weller]

Hi Palash,

If you are handling your own routing, a Private Gateway will be a lot more
performant.

-Si

On Tue, Feb 27, 2024 at 7:09 AM Palash Biswas  wrote:

> Hi everyone,
>
> I'm seeking advice on the best method to connect VPCs across different
> accounts. Should I use a private gateway or a site-to-site VPN?
>
> I have two accounts and want to establish communication between the VPC in
> Account A and the VPC in Account B.
>
> Thanks,
> Palash Biswas
>

delete old SystemVM Templates

[me]

Hello,

is it possible to delete old, not used anymore, SystemVM Templates via UI?
And does it make sense to delete them?

After upgrading from 4.18.1.0 to 4.19.0.0 we do have 2 systemvm templates
registered (as expected) in our CS. The first one is called "SystemVM
Template (KVM)" and the other one is called "systemvm-kvm-4.19.0".

router.template.kvm points to the new systemvm-kvm-4.19.0 one. SystemVM
Template (KVM) is not being used by any instance and/or router. When I try
to delete it via UI, I get: "Error The DomR template cannot be deleted." I
did not try to force it.

Should you cleaup your templates or do you just leave it there?

 

Regards,

Swen

Re: Advice Needed: Connecting VPCs Across Different Account

[Wei ZHOU]

private gateways with the same vlan is better.
However, you need an admin user to configure the private gateways.

-Wei

On Tue, Feb 27, 2024 at 2:09 PM Palash Biswas  wrote:

> Hi everyone,
>
> I'm seeking advice on the best method to connect VPCs across different
> accounts. Should I use a private gateway or a site-to-site VPN?
>
> I have two accounts and want to establish communication between the VPC in
> Account A and the VPC in Account B.
>
> Thanks,
> Palash Biswas
>

Re: delete old SystemVM Templates

[Wei ZHOU]

if it is not needed, of course you can remove it
- update template type to USER
- delete the template from zones


-Wei

On Tue, Feb 27, 2024 at 4:56 PM  wrote:

> Hello,
>
> is it possible to delete old, not used anymore, SystemVM Templates via UI?
> And does it make sense to delete them?
>
> After upgrading from 4.18.1.0 to 4.19.0.0 we do have 2 systemvm templates
> registered (as expected) in our CS. The first one is called "SystemVM
> Template (KVM)" and the other one is called "systemvm-kvm-4.19.0".
>
> router.template.kvm points to the new systemvm-kvm-4.19.0 one. SystemVM
> Template (KVM) is not being used by any instance and/or router. When I try
> to delete it via UI, I get: "Error The DomR template cannot be deleted." I
> did not try to force it.
>
> Should you cleaup your templates or do you just leave it there?
>
>
>
> Regards,
>
> Swen
>
>

Re: delete old SystemVM Templates

[Wei ZHOU]

No, you are free to remove any template which is not used any more.
for SYSTEM templates, there is a check to guard the templates to be removed
by mistake. so you need to change the type to USER then remove it.

-Wei

On Tue, Feb 27, 2024 at 5:04 PM  wrote:

> Thx! Is there any reason to keep old templates when no instance is using
> it anymore?
>
> -Ursprüngliche Nachricht-
> Von: Wei ZHOU 
> Gesendet: Dienstag, 27. Februar 2024 16:58
> An: users@cloudstack.apache.org
> Betreff: Re: delete old SystemVM Templates
>
> if it is not needed, of course you can remove it
> - update template type to USER
> - delete the template from zones
>
>
> -Wei
>
> On Tue, Feb 27, 2024 at 4:56 PM  wrote:
>
> > Hello,
> >
> > is it possible to delete old, not used anymore, SystemVM Templates via
> UI?
> > And does it make sense to delete them?
> >
> > After upgrading from 4.18.1.0 to 4.19.0.0 we do have 2 systemvm
> > templates registered (as expected) in our CS. The first one is called
> > "SystemVM Template (KVM)" and the other one is called
> "systemvm-kvm-4.19.0".
> >
> > router.template.kvm points to the new systemvm-kvm-4.19.0 one.
> > SystemVM Template (KVM) is not being used by any instance and/or
> > router. When I try to delete it via UI, I get: "Error The DomR
> > template cannot be deleted." I did not try to force it.
> >
> > Should you cleaup your templates or do you just leave it there?
> >
> >
> >
> > Regards,
> >
> > Swen
> >
> >
>
>
>

Re: CKS Storage Provisioner Info

[Wei ZHOU]

Hi,

The CloudStack management server must be accessible from the pod
"cloud-controller-manager" in the CKS cluster.

It looks like the issue is caused by your network configuration, not
CloudStack.

-Wei

On Wed, Feb 28, 2024 at 7:19 AM Bharat Bhushan Saini
 wrote:

> Hi Kiran,
>
>
>
> I used a private ip range which is bridged in my internal organisation
> network.
>
>
>
> JUST FYI, I also tested a simple deployment but it is also didn’t took
> external ip.
>
>
>
>
>
> Thanks and Regards,
>
> Bharat Saini
>
>
>

Re: CKS Storage Provisioner Info

[Jayanth Babu A]

Hello Bharat,
Please confirm if you’re using self-signed or a signed certificate from trusted 
CA on your management server API endpoint. The error indicates that the CKS 
cannot complete the TLS handshake due to the common name or SAN mismatch.
It is usually advised to put Trusted CA signed-certificates and have the proper 
domain names or SAN that match the certificate. Please change the endpoint 
domain name accordingly in the global configuration setting “endpoint.url” and 
ensure the CKS nodes or network is able to resolve and connect to it.

Thanks,
Jayanth

From: Bharat Bhushan Saini 
Date: Wednesday, 28 February 2024 at 10:19
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
Hi Jayanth,

I shows below

Error syncing load balancer: failed to ensure load balancer: error retrieving 
load balancer rules: Get 
https://checkpoint.url-protection.com/v1/url?o=https%3A//10.x.x.2%3A8443/client/api%3FapiKey%3Dk83H56KFdhFqpv7cXPU11nkwxPt8f2rXnm1WWVIRdeErqZr72Pzp7ySmricPWs7FQQuMmClznDhMz7uqnRD2wA%26amp%3Bcommand%3DlistLoadBalancerRules%26amp%3Bkeyword%3Da85de8a8ca0f543cca49e06fc3f30791%26amp%3Blistall%3Dtrue%26amp%3Bresponse%3Djson%26amp%3Bsignature%3DHVQNyOboMkqp4rwQYY6wLJ1NBzY%253D%3A=ZDM3OWY0MmUwMWVmMGVlMA===YTJhNGFiMWNiNzY1YWE1YWYzOTdhMzk0ZjlkNDkwYTgyMDNjOGQ5ZDU0ODM4MDcxMjAyNjQzOGM4MGI3NTFhZg===Y3AxZTpueHRnZW5pbmZpbml0ZWRhdGFjZW50ZXI6YzpvOjNiNDk1M2JiNmRjNDE0OGZlYmU4NTY0YmUzYjA4ODU4OnYxOnQ6VA==
 x509: cannot validate certificate for 10.x.x.2 because it doesn't contain any 
IP SANs

Can I put API URL as management URL?

Thanks and Regards,
Bharat Saini

[signature_1763564491]

From: Jayanth Babu A 
Date: Wednesday, 28 February 2024 at 10:15 AM
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
EXTERNAL EMAIL: Please verify the sender email address before taking any 
action, replying, clicking any link or opening any attachment.


Hello Bharat,
What does “# kubectl describe svc/traefik” say?

Thanks,
Jayanth

From: Bharat Bhushan Saini 
Date: Wednesday, 28 February 2024 at 10:11
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
Hi All,

I had some queries regarding CKS service

Why external IP is not shown in SVC. Do we have any resolvent for this and 
traefik service is in pending state as I have 1 control and 1 worker node.

kubectl get svc

NAME TYPE   CLUSTER-IP   EXTERNAL-IP   PORT(S)  
AGE

kubernetes   ClusterIP  10.96.0.1443/TCP  
39h

nginx-nodeport   NodePort   10.110.79.87 80:30918/TCP 
15h

nginx-serviceNodePort   10.110.244.170   80:32180/TCP 
17h

traefik  LoadBalancer   10.107.145.95 
80:32289/TCP,443:32426/TCP   17h

Inside the cluster the CLUSTER IP was continuous change by the 
SVC, I just login again and I saw that the CLUSTER IP is different from 
yesterday.

Thanks and Regards,
Bharat Saini

[signature_3834506047]

From: Jayanth Babu A 
Date: Tuesday, 27 February 2024 at 11:26 AM
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
EXTERNAL EMAIL: Please verify the sender email address before taking any 
action, replying, clicking any link or opening any attachment.


Hello Bharat,
You’ve some well-known options as below:
https://github.com/kubernetes-sigs/nfs-subdir-external-provisioner

Re: CKS Storage Provisioner Info

[Jayanth Babu A]

Hello Bharat,
What does “# kubectl describe svc/traefik” say?

Thanks,
Jayanth

From: Bharat Bhushan Saini 
Date: Wednesday, 28 February 2024 at 10:11
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
Hi All,

I had some queries regarding CKS service

Why external IP is not shown in SVC. Do we have any resolvent for this and 
traefik service is in pending state as I have 1 control and 1 worker node.

kubectl get svc

NAME TYPE   CLUSTER-IP   EXTERNAL-IP   PORT(S)  
AGE

kubernetes   ClusterIP  10.96.0.1443/TCP  
39h

nginx-nodeport   NodePort   10.110.79.87 80:30918/TCP 
15h

nginx-serviceNodePort   10.110.244.170   80:32180/TCP 
17h

traefik  LoadBalancer   10.107.145.95 
80:32289/TCP,443:32426/TCP   17h

Inside the cluster the CLUSTER IP was continuous change by the 
SVC, I just login again and I saw that the CLUSTER IP is different from 
yesterday.

Thanks and Regards,
Bharat Saini

[signature_3834506047]

From: Jayanth Babu A 
Date: Tuesday, 27 February 2024 at 11:26 AM
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
EXTERNAL EMAIL: Please verify the sender email address before taking any 
action, replying, clicking any link or opening any attachment.


Hello Bharat,
You’ve some well-known options as below:
https://github.com/kubernetes-sigs/nfs-subdir-external-provisioner
https://github.com/rook/rook
 - CephFS & NFS can fit

Thanks,
Jayanth Reddy

From: Bharat Bhushan Saini 
Date: Tuesday, 27 February 2024 at 11:01
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
Hi All,

Thanks for helping out for the storage provisioner.

In CKS service I am able to bound PVC as ReadWriteOnce with Leaseweb CloudStack 
CSI Driver.
Is there any opportunity in the cloudstack with other driver that I can bound 
storage as ReadWriteMany.

I tried with Leaseweb and got below warning
Warning  ProvisioningFailed6s (x5 over 21s)  
csi.cloudstack.apache.org_cloudstack-csi-controller-7f89c8cd47-ztllw_c6112933-3587-4445-8702-74b255b1e56f
  failed to provision volume with StorageClass "cloudstack-custom": rpc error: 
code = InvalidArgument desc = Volume capabilities not supported. Only 
SINGLE_NODE_WRITER supported.

Thanks and Regards,
Bharat Saini

[signature_636195584]

From: Kiran Chavala 
Date: Monday, 26 February 2024 at 5:54 PM
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
EXTERNAL EMAIL: Please verify the sender email address before taking any 
action, replying, clicking any link or opening any attachment.


Hi Bharath

Note the CKS provisioner works on KVM based cloudstack environment

Regards
Kiran

From: Kiran Chavala 
Date: Monday, 26 February 2024 at 5:45 PM
To: users@cloudstack.apache.org 
Subject: Re: CKS Storage Provisioner Info
Hi Bharat Bhusan

Please follow these steps


1. Deploy a Kubernetes cluster on cluster


NAME STATUS   ROLES   AGE VERSION
ty-control-18de52c04f7   Readycontrol-plane   5m11s   v1.28.4
ty-node-18de52c4185  Ready  4m55s   v1.28.4

kubectl get secrets -A
NAMESPACE  NAME  TYPE   
 DATA   AGE
kube-systemcloudstack-secret Opaque 
 1  10m



2. Check the deployment, it should  be in pending

~ kubectl get deployments -A
NAMESPACE  NAMEREADY   UP-TO-DATE   
AVAILABLE   AGE
kube-systemcloudstack-csi-controller   0/1 10   
46s


3. Edit the deployment and remove the nodeSelector part

~kubectl edit deployment/cloudstack-csi-controller -n kube-system


  nodeSelector:
kubernetes.io/os: linux
node-role.kubernetes.io/master: ""

4. Check the deployment again  and it should be running state

~kubectl get deployments -A
NAMESPACE  NAMEREADY   UP-TO-DATE   
AVAILABLE   AGE
kube-systemcloudstack-csi-controller   1/1 11   
2m39s

5. Replace the disk offering in the storage class yaml

Provide the custom disk offering UUID (service offerings > disk offering > 
custom disk 

Cloudstack + Supabase Anyone?

[Hunter Yap]

Hi Guys,

I recently came upon Supabase Open Source which has managed Postgres DB.

Wondering if anyone uses them? Or have integrated it with Cloudstack to
provided Managed DB before?

Regards,
Hunter

AW: delete old SystemVM Templates

[me]

Thx! Is there any reason to keep old templates when no instance is using it 
anymore?

-Ursprüngliche Nachricht-
Von: Wei ZHOU  
Gesendet: Dienstag, 27. Februar 2024 16:58
An: users@cloudstack.apache.org
Betreff: Re: delete old SystemVM Templates

if it is not needed, of course you can remove it
- update template type to USER
- delete the template from zones


-Wei

On Tue, Feb 27, 2024 at 4:56 PM  wrote:

> Hello,
>
> is it possible to delete old, not used anymore, SystemVM Templates via UI?
> And does it make sense to delete them?
>
> After upgrading from 4.18.1.0 to 4.19.0.0 we do have 2 systemvm 
> templates registered (as expected) in our CS. The first one is called 
> "SystemVM Template (KVM)" and the other one is called "systemvm-kvm-4.19.0".
>
> router.template.kvm points to the new systemvm-kvm-4.19.0 one. 
> SystemVM Template (KVM) is not being used by any instance and/or 
> router. When I try to delete it via UI, I get: "Error The DomR 
> template cannot be deleted." I did not try to force it.
>
> Should you cleaup your templates or do you just leave it there?
>
>
>
> Regards,
>
> Swen
>
>