Troubleshooting CCMv2

This page lists common issues related to troubleshooting workload connectivity using CCMv2.

Incorrect network setup


Reason for connection failure	How to fix the problem
For CCM to work, the agent must be able to connect to its counterpart running inside the Control Plane. A misconfigured network often blocks this path. This can happen in following ways: Customer's network does not resolve CCM DNS addresses. Customer's network does not allow outbound connections. Customer's network does not allow outbound connections to Cloudera IPs. Customer's network does not allow outbound connections to the required ports. Customer expects us to use a non-transparent proxy which does not allow connection to the required Cloudera IPs / ports. Customer expects us to use a non-transparent proxy which does deep-packet inspection to sniff underlying traffic and blocks CCM traffic.	Running the `cdp-telemetry doctor` command helps diagnose most problems of this category. See Collecting workload logs. Fixing the problem usually means ensuring that you set up the outbound network access destinations that are listed as CDP prerequisites. See: AWS outbound network access destinations Azure outbound network access destinations GCP outbound network access destinations

The easiest and recommended way to collect logs is to use the cdp-telemetry doctor tool.

The ccm module of the cdp-telemetry doctor tools collects all the information required on a given node.

cdp-telemetry doctor ccm status
cdp-telemetry doctor ccm report

To generate the diagnostic bundle, you need to ssh into each node of concern and invoke the following command:

cdp-telemetry doctor ccm report