Here's another interesting issue I came across the other day that I thought I would share with you just in case you happen to find yourself in a similar situation. Hopefully if you do then this will help you get all this working the way it should.
Scenario: An RMS in a parent domain and client agents that are domain controllers in a child domain in a DMZ.
The manual agent install goes fine on the clients but the agents never appear in the operators console despite Review New Manual agent installations in pending management View and Auto–approve New manually installed agents settings in SETTINGS—SECURITY –GENERAL.
The following event shows up on the agents:
Event Type: Error Event Source: OpsMgr Connector Event Category: None Event ID: 20070 Computer: DC Description: The OpsMgr Connector connected to <domain>, but the connection was closed immediately after authentication occurred. The most likely cause of this error is that the agent is not authorized to communicate with the server, or the server has not received configuration. Check the event log on the server for the presence of 20000 events, indicating that agents which are not approved are attempting to connect.
This event shows up on the server:
Event Type: Error Event Source: OpsMgr Connector Event Category: None Event ID: 20002 Description: A device at IP <addr> attempted to connect but could not be authenticated, and was rejected.
None of the agents show up in any of the following tables under Opsmgrdb:
Dbo.Mt_Computer Dbo.Mt_healthservice Dbo.Mt_healthservicewatcher dbo.AgentPendingActions
The following powershell command returns nothing:
get-agentpendingaction
The product documentation does not talk much about this scenario other than having port 5723 open from the agent to the server:
http://technet.microsoft.com/da-dk/library/cc540431(en-us).aspx
Regardless, what I’ve found is that we also need to have port 88 and port 389 opened between the agent and the RMS if they’re separated by a firewall. This has worked for me just about every time I’ve found myself in this situation.
Hope this helps,
Rohit Kaul
Feed: The Operations Manager Support Team Blog Posted on: Tuesday, February 17, 2009 10:38 AM Author
Feed: The Operations Manager Support Team Blog Here's another interesting issue I came across the
Hi,What does tcp port 88 used for?
I have another problem about the firewall.
the RMS and the clients are in the same domain but they were separated by the isa 2006.I have opened the protocols such as to join a computer to the domain and tcp port 5723.and now I am use an action account as a member of a domain admins,but i can't install the agents to the client because of the rpc unreachable.I try to open tcp port 135 and 1024 plus ports,but no results ,What another ports should I opened?
Check this one out here
========================
Agent push requirements (including firewall ports):
The account being used to push the agent must have local admin rights on the targeted agent machine.
The following ports must be open:
RPC endpoint mapper Port number: 135 Protocol: TCP/UDP
*RPC/DCOM High ports (2000/2003 OS) Ports 1024-5000 Protocol: TCP/UDP
*RPC/DCOM High ports (2008 OS) Ports 49152-65535 Protocol: TCP/UDP
NetBIOS name service Port number: 137 Protocol: TCP/UDP
NetBIOS session service Port number: 139 Protocol: TCP/UDP
SMB over IP Port number: 445 Protocol: TCP
MOM Channel Port number: 5723 Protocol: TCP/UDP
The following services must be set:
Display Name: Netlogon Started Auto Running
**Display Name: Remote Registry Started Auto Running
Display Name: Windows Installer Started Manual Running
Display Name: Automatic Updates Started Auto Running
Extracted from : http://blogs.technet.com/kevinholman/archive/2007/12/12/agent-discovery-and-push-troubleshooting-in-opsmgr-2007.aspx
In my own experience, on dmz environments, no need for those 2 ports (88 and 389); but ensure that the RMS server is able to resolve from IP to FQDN. a problem with the DNS configuration will yield event IDs 20002, and all the agents on grey state.