Further explanation for anyone reading this thread...
CIP paths are always made up from pairs of specification values, viz...
Port, Destination, [Port, Destination], [Port Destination], and so on.
So a proper CIP path should consist of an even number of specifications, separated by commas - but read on....
To demonstrate, I will split up the OP's corrected path (courtesy Ken Roach) into its constituent parts.
LOCAL_AREA_NETWORK, 2, 128.1.49.210, 1, 6 ,2 ,128.1.50.240 ,1 ,0
LOCAL_AREA_NETWORK
This is already "defined" as 1,1 in the I/O config - Port = 1, the Backplane port : Destination = Slot 1, the local Ethernet module in the L55 chassis.
2, 128.1.49.210
Port = 2, the "Front" port (the RJ45 Ethernet port) : Destination = IP Address 128.1.49.210, the first Ethernet module in the L73 chassis.
1, 6
Port = 1, the Backplane port : Destination = Slot 6, the second Ethernet module in the L73 chassis.
2 ,128.1.50.240
Port = 2, the "Front" port (the RJ45 Ethernet port) : Destination = IP Address 128.1.50.240, the Ethernet address of the CompactLogix.
1 ,0
Port = 1, the Backplane port : Destination = Slot 0, the CompactLogix processor (always slot 0 for a CompactLogix).
You can see we have 9 parts to this path-spec, which is not an even number !
The start of the path could have been entered explicitly as ...
1,1, 2,128.1.49.210, 1, 6 ,2 ,128.1.50.240 ,1 ,0
But the system will automatically replace anything it "knows about" (from the I/O configuration), with the assigned name. In this case 1,1 is known in the I/O Config as LOCAL_AREA_NETWORK. This sort of hides the "pairing" specification from the unwary. A single "name" (an odd number) will replace an even number of path specification values.
I always teach students to construct CIP paths explicitly, and then they see the names appear as if by magic. Then when they have the idea firmly in their grasp, they can then use assigned I/O configuration names instead of a pair of CIP path specification values.