Duplicate key value violation while applying the compute profile with TKGI clusters

Products

• VMware Tanzu Kubernetes Grid

Issue/Introduction

Symptoms:

• Once applying the  tkgi compute-profile tasks fails with a bosh timeout, the subsequent deploys may fail with database error.

• bosh task shows an error similar to:
  Task 6840762 | 07:48:56 | Error: PG::UniqueViolation: ERROR:  duplicate key value violates unique constraint "link_providers_constraint"
  DETAIL:  Key (deployment_id, instance_group, name, type)=(32, worker-large-dev, docker, job) already exists.

Environment

VMware Tanzu Kubernetes Grid Integrated Edition 1.x

Cause

If a failure occurs when you apply the compute profile ( can be because of renaming the instances  or unrelated timeouts such as once agent, once a unmount disk) the  subsequent deploys would also fails till we fix the the first failure.

Resolution

This is a known issue and currently there is no resolution.

Workaround:

To work around this issue:

1. find the deployment ID in the logs (is part of the error message):
   {"time":1632124136,"error":{"code":100,"message":"PG::UniqueViolation: ERROR:  duplicate key value violates unique constraint \"link_providers_constraint\"\nDETAIL:  Key (deployment_id, instance_group, name, type)=(32, worker-large-dev, docker, job) already exists.\n"}}', "result_output" = '', "context_id" = '225d2ac9-9415-4de4-ba7e-99696ce23484' WHERE ("id" = 6840762)
   
   From above  Key (deployment_id, instance_group, name, type)=(32, worker-large-dev, docker, job)  so deployment_id is 32.

2. Next ssh into bosh director and become root sudo su –

3. Start the director console: /var/vcap/jobs/director/bin/console

4. Execute the following snippet in the console where 32 should be replace with the actual deployment id found in step 1:
   BD::Models::Links::LinkProvider.where(deployment_id: 32).map(&:delete)
   BD::Models::Links::LinkConsumer.where(deployment_id: 32).map(&:delete)
   
   exit

5. Get a copy of the manifest used from the last failed deploy.
   bosh task {task_id_of_the_failed_deploy} --debug | awk '/Manifest:/{f=1;next} /D,|I,/{f=0} f' | bosh int - > manifest.yml

6. Perform a normal bosh deploy with the retrieved manifest. 
   bosh deploy -d service-instance_9d44b100-2f34-4ee5-9d96-a9510fec7a55 manifest.yml