ThinLinX Support > ThinLinX Help Desk > Knowledgebase

Search help:

How do client devices find the TMS server?


The TMS agent's (tms_client's) default TMS server discovery behaviour (the "auto" method) tries the "dhcp" method first, then the "static" method (values default to hostname "tms" and port 8085), and finally the "broadcast" method.

The "dhcp" method will look for custom DHCP options (231 and 232) in the response from your DHCP server that specify a TMS server address.  See Chapter 7 of the TMS User Manual ( for more information on how to set these options.

The "static" method will look for TMS at a fixed address.  By default, this is tms:8085, so if you add a DNS CNAME or A record named "tms" that points at the canonical hostname (CNAME) or IP addresss (A) of your TMS server in a domain that's included in your clients' search path, then your devices will be able to locate your server using default values.  You could of course manually specify the TMS server's real hostname/IP address and TCP port number instead, but since the devices presumably have no access to TMS yet, that would have to be done locally on each device using the local configuration tool (Tlxconfig).

The last resort "broadcast" method listens for broadcasts from TMS on UDP port 9097 in which TMS provides address details.  This will only work if the client is on the same IP subnet (or possibly specific Ethernet segments if your switch is configured not to pass broadcast packets for any reason, e.g. when bridging) as the TMS server.  If your TLXOS devices are on a different IP subnet than your TMS server, then you will have to provide them with a server location hint via the "static" or "dhcp" methods.

Starting from TLXOS 4.10.0 and 4.8.2 LTS, we added the TMS beacon as a TMS discovery aid.  This is a TMS broadcast repeater, which a nominated TLXOS client on a different subnet than the TMS server can be configured to start.  Clients on the same IP subnet as the beacon that use (or fall through to) the broadcast discovery method will detect this signal and connect to the TMS server that it names.  This will of course stop working if the client that runs the beacon is shut down or put into Maintenance Mode.

Once a client is connected to a TMS server, it will keep the connection open and talk to that server exclusively until it is rebooted (sending keepalives every few minutes to prevent intervening firewalls from closing the connection).  If clients can fall through to UDP discovery, it is therefore important that you do not run multiple copies of TMS in the same IP subnet, or else clients will connect to whichever TMS instance they detect first and will not appear in the other.

Related articles Why aren't my devices showing up in TMS?
How can I protect clients from rogue/unauthorized TMS servers?
Can I install/upgrade/run TLXOS via PXE boot?
Can TMS be used to manage remote devices connecting via the Internet?
Article details
Article ID: 10
Category: Frequently Asked Questions
Date added: 2019-05-29 00:52:00
Views: 970

« Go back