Transport Types

Each transport is a separate communication medium, insulated against crosstalk from other transports.

Using TIBCO FTL, you can use a combination of the following transport types to meet your organization's communication needs.

•

Dynamic TCP (DTCP): TCP-based connectivity without the need to specify explicit host and port configurations

•

Static TCP: TCP-based connectivity using specific host and port configurations

•

Auto: TCP connectivity to the FTL Server using only a transport name

•

Multicast: UDP-based message delivery supporting potentially very large numbers of subscribers (requires multicast support to be enabled within the network)

•

Process: High-speed connectivity within a single process

•

Shared Memory: High-speed connectivity between programs running on the same host computer

•

Direct Shared Memory: High-speed communication between programs running on same host computer (highest performance available with some API constraints)

•

Reliable UDP (RUDP): UDP-based datagrams connectivity (reliability of TCP with more user control and configuration options)

In the following diagram, a publisher and subscribers that are co-located on the same host use a shared memory transport (T1). The subscriber using a multicast transport (T2) is located on a separate host, local area network (LAN), or wide area network (WAN). The publisher sends a message once and TIBCO FTL transfers the message over both transports. The two transports carry identical message streams and the endpoints (EP1) have the same transport details.