AIOps 21.3.1 - Unable to login to newly created Tenants, the login process enter in a loop if the top domain name is ".local"

Products

• DX Operational Intelligence
• DX Application Performance Management
• CA App Experience Analytics

Issue/Introduction

We can connect as MASTERADMIN but not to newly created Tenants

We can see the call to the dxiportal to cloudmanagement portal and this process keeps repeating over and over.

 

Cause

 

From Developer Tools > Application, we can see that CA_CLOUD_MANAGEMENT cookie cannot be created, the CA_CLOUD_MANAGEMENT cookie was created at the subdomain instead of the their base domain ".local". This cookie was supposed to be shared between all the subdomains so it had to be created at the base level. The root cause is because of the top domain ".local" is not handled property, this is problem is related to defect DE522893 

Environment

DX Operational Intelligence 21.3.1
DX Application Performance Management 21.3.1
DX AXA 21.3.1

Resolution

You can find below the steps how to apply the hotfix for DX Platform 21.3.1 ONLY

 

1. Download and copy attached image to the kubernetes/openshift master server

2. Load a new image to the local docker registry

docker load < manager:21.7.0.27-linux-amd64.tar

Verify image has been loaded:

docker images 

For example:

docker images | grep manager

3. Tag the imported image:

docker image tag oerth-scx.lvn.broadcom.net:8443/apmservices/manager:<version>-linux-amd64 <private-registry>/dxi/apmservices-manager:<version>-linux-amd64

Change <private-registry> with your registry endpoint, for example:

docker image tag oerth-scx.lvn.broadcom.net:8443/apmservices/manager:21.7.0.27-linux-amd64 munqa001499.bpc.broadcom.net:5000/dxi/apmservices-manager:21.7.0.27-linux-amd64

Verify image tag has been updated:

docker images 

For example:

docker images | grep manager

3.  Push the image to the private registry:

docker image push <private-registry>/dxi/apmservices-manager:21.7.0.27-linux-amd64

for example:

docker image push munqa001499.bpc.broadcom.net:5000/dxi/apmservices-manager:21.7.0.27-linux-amd64

5e1a592aa0a6: Pushed
00d7593bf56a: Pushed
1adfe9cb233d: Pushed
a1bbdbc60c33: Pushed
01c81433d847: Pushed
7184f5f9c79e: Pushed
cda81c0c4540: Pushed
e2eb06d8af82: Mounted from dxi/doi-21.3.1.ga-servicestatemanager
21.7.0.27-linux-amd64: digest: sha256:183ea4ea3200c983827021303e0e268f4847985fdcc5b3456aaba507649089d8 size: 2009

 

4. Edit the apmservices-manager-001 deployment :

kubectl edit deployment apmservices-manager-001 -n<namespace>

Locate the containers image section, like below:

Current value:

containers:
      image: <private-registry>/dxi/apmservices-manager:21.7.0.25-linux-amd64       

Update the image name, for example:

          image: munqa001499.bpc.broadcom.net:5000/dxi/apmservices-manager:21.7.0.27-linux-amd64

Save the file,  apmservices-manager-001pod will be restarted 

 

5. Verify pod is using the new image:

kubectl describe po <apmservices-manager-pod> -n<namespace> | grep Image

For example:

      

Additional Information

https://knowledge.broadcom.com/external/article/190815/dx-aiops-troubleshooting-common-issues-a.html