We have a server on the subnet set as the local relay for all workstations on the same subnet.
In the BES Console, the PCs are showing as reporting in correctly, ie Last Report Time retrieved property is showing times/dates as we would expect but for some reason deployments aren’t being picked up by the machines.
Any advice on tracking down the breakdown in communication ? Our network guys have said that there shouldn’t be a firewall issue.
Sounds like you are saying the communication is good from the BES Clients via the Relay to the BES Server - hence the successful property reports.
On the other hand, the communication from the BES Server to the BES Clients to notify them of Actions (Deployments) are not being received.
From the BES Clients to the BES Server, BigFix uses one TCP port of choice (default is 52311). Looks like this one is working.
From the BES Server to the BES Clients, BigFix uses one UPD port of choice (same as the above, or the default is 52311). Looks like this one is NOT working. This UDP notification is used simply to ‘tickle’ the BES Clients into checking back with the Server to retrieve new instructions (such as your deployment action). If this is blocked, the BES Clients will be default still check in after 24 hours.
Read traffic A for Client to Server communication.
Read traffic B for Server to Client UDP notification.
One easy way to test the UDP ping is to right-click in the BES Console on the Client with the problem, then select “Send Refresh”.
On the BES Client side, check the log file (example location: C:\Program Files\BigFix Enterprise\BES Client__BESData__Global\Logs\20080929.log) for the line that starts with “ForceRefresh command received…”. If you don’t see the line with less than 10 seconds, the UDP message has not gotten through to the Client.
To modify how often the BES Clients checks in with the BES Server (default 24 hours), read article, “How do I change the gather interval for a BES Client (how to enable command polling)?” at http://support.bigfix.com/cgi-bin/kbdirect.pl?id=185.
If so, my guess is that your clients are still talking to the old relay. The old relay probably can’t get UDP notifications through to the new IP range, but clients can get TCP out to the old relay.
Clients will auto relay select to the new relay at some point. When depends on your deployment settings but it’s probably every 24 hours. My guess is you’ll see this issue resolve itself once all the machines have gone through an automatic relay selection cycle in their new location.
Right, I’ve tested the right click->Send Refresh and there was nothing in the client’s logfile so I’m assuming the UDP packets are not reaching their destination. I’ve asked our network admins to take a look.