Intel® Edge Software Hub
Get answers from community peers to your questions about building Edge Software Hub solutions for edge compute node.
Announcements
Welcome to the Intel Community! If you like the answer to your question, please mark it as 'Accepted Solution' to help others.

The Edge Software Vision Package for Red Hat Enterprise Linux is now available here.

Learn more about developing on Intel® Hardware and Software here.
394 Discussions

Kubernetes cluster deployment configuration for Intelligent Traffic Management Version: 2022.1

louisdace
Novice
572 Views

Hi Team,

Can you help with Kubernetes cluster deployment configuration for Intelligent Traffic Management Version: 2022.1 https://www.intel.com/content/www/us/en/developer/articles/reference-implementation/intelligent-traffic-management.html .

Deployments cannot be scheduled on control plane nodes. Currently the rule-engine, itm-analytics, influxdb, cloud-connect, and item-dashboard deployments are all scheduled on the master nodes. Can this be changed to run on worker nodes without impact? 

HostPath volume mounts are not allowed. Within our deployment there are statically allocated PVs available which can be assigned to workloads using PersistentVolumeClaims. Can this be changed in the itm-video-inference pod?

The itm-video-inference has a privileged pod. Can you share what specific capabilities are needed or if the pod can run as a non-privileged pod? 

 

 

 

0 Kudos
1 Solution
JesusE_Intel
Moderator
468 Views

Hi louisdace,

 

I'm sorry for the delay in our response, I reached out to the development team for more help.

 

The rule-engine, itm-analytics, influxdb, cloud-connect, and item-dashboard deployments can be changed to run on worker nodes without impact. Just need to change the nodeSelector in the yaml file. For example: nodeSelector: label-test:"worker", while you should also label the nodes, such as kubectl label nodes workernode label-test=worker.

s1.png

To change the hostpath in the itm-video-inference pod, you need to change three places.1) change values.yaml, add volumes , 2) add itm-video-inference-pv.yaml in services/templates, 3) change hostpath in itm-video-inference.yaml

s2.png

The itm-video-inference pod can run as a non-privileged pod, just comment this code in the itm-video-inference.yaml

securityContext:

#privileged: true

readOnlyRootFilesystem: true

 

Hope this helps!

 

Regards,

Jesus

 

 

 

View solution in original post

0 Kudos
3 Replies
JesusE_Intel
Moderator
547 Views

Hi louisdace,


Thanks for reaching out, let me run a quick tests on my end and get back to you with my findings.


Regards,

Jesus


0 Kudos
JesusE_Intel
Moderator
469 Views

Hi louisdace,

 

I'm sorry for the delay in our response, I reached out to the development team for more help.

 

The rule-engine, itm-analytics, influxdb, cloud-connect, and item-dashboard deployments can be changed to run on worker nodes without impact. Just need to change the nodeSelector in the yaml file. For example: nodeSelector: label-test:"worker", while you should also label the nodes, such as kubectl label nodes workernode label-test=worker.

s1.png

To change the hostpath in the itm-video-inference pod, you need to change three places.1) change values.yaml, add volumes , 2) add itm-video-inference-pv.yaml in services/templates, 3) change hostpath in itm-video-inference.yaml

s2.png

The itm-video-inference pod can run as a non-privileged pod, just comment this code in the itm-video-inference.yaml

securityContext:

#privileged: true

readOnlyRootFilesystem: true

 

Hope this helps!

 

Regards,

Jesus

 

 

 

0 Kudos
JesusE_Intel
Moderator
429 Views

If you need any additional information, please submit a new question as this thread will no longer be monitored.


0 Kudos
Reply