Recently we are installing client in a IEM shared environment.
Customers will not connect to the TEM server directly only have connections with their respective relays.
We observed that after connecting clients properly, time after, they lose connection and fail to report.
Initially when we connect the client, it has Relay connection in manual option, and it should configure his two relays.
this step is done, but after time (not known how many time) this client lose his configuration into the besclient.config file and the relay servers desapear.
When this happens, do the clients also move into Automatic Relay Selection mode? Check the registry for client settings relating to relay selection.
I’ve seen this happen when we had a Policy Action to switch clients into Automatic relay selection mode, when the clients are unable to detect any relays due to blocked ICMP traffic to the relays.
In that case, we can either make an exception to leave these clients in Manual mode, or you can fudge the clients’ HOSTS file to make the root server hostname map to the IP address of the desired relay.
You can also instruct relays to report their names and IP’s by using the following setting on the relays that you are having issues with
_BESClient_Relay_NameOverride
Type: Boolean
Version: 6.0
Platform: All
Default:
Requires Client Restart: NO
Description: Affects the RelayNameOfClient. Often set to IP of machine so in instances where DNS cannot resolve a relay's name the IP can be selected.
This will make the relays.dat have the IP number in it as well so machines that cannot resolve the name can try the IP. This requires the relay to report and the clients trying to use it to gather an actionsite with the relays.dat in it so it can take a bit to actually give help to your clients.
About version, really yes, is the same.
About clients connection, we needed to add some checks to verify we have correct conectivity btw relays <–> clients.
In some cases has been detected that problem, but in the most of them we are rigth.
The last step I did was to reinstall the agent en all the clients I found with the problem and after that I loaded some performance settings checking into the DB, really all the lines of the fixlet were executed without problems.
By the moment it seems reinstall action took effect. But I have no more information