The larger the DVO metrics are, the more time it takes for the Insights Operator to read it. Previously we added retry logic to the Inisghts Operator DVO gatherer to retry few times if the DVO service exists, but doesn't return HTTP 200 or connection is refused. The logic was not very good, because it used Go context with 5 seconds timeout. 

1. Deploy DVO to the cluster
2. Create few namespaces introducing some DVO failing checks (the more lines in the DVO metrics, the better)
3. Run Insights Operator to gather the DVO metrics

Insights Operator doesn't gather any DVO data and the gatherer reports "context canceled"

Insights Operator gathers complete DVO data