In an environment I was working in recently we had several clients which had gone grey and would not communicate with OpsMgr. We were unable to push the agent from the OpsMgr console due to firewall restrictions. We were also unable to get the agent to communicate with a Management Server or a Gateway. Most of the time we can debug this through telnet to validate effective name resolution and communication on port 5723. However while they could name resolve and could communicate on 5723 they still would not update their status in OpsMgr. The following are the steps which we ended up taking to resolve the issue:

  1. Remove the agent from the server in add/remove programs
  2. Delete the agent in the Operations Manager console
  3. Reinstall the agent manually
  4. Re-open the Operations Manager console with the /clearcache switch if it’s not listed
  5. Approve the manually installed agent
  6. Restart the agent on the server
  7. Verify successful management pack deployment to the agent using the OpsMgr log
  8. Verify healthy status in the Operations Manager console – agent managed section