Output groups on the network queues are grouped by data set forms requirements, SYSOUT class, and other selection criteria. If you decide to route output on the network queue to the local node, remember that routing output does involve calls to an installation's security product, which can impact command performance. JES2 removes the output group from the network queue and places it on the appropriate local class queue. Similarly, JES2 routes output groups originally destined for local output to another node by removing them from their class queue, changing their route codes, and placing them on the network queue.

The JES2 commands that deal with output routing (for example, $R and $O) allow routing and rerouting of data sets. JES2 places the current routing of a data set in the data set header whenever the data set is transmitted.

The network SYSOUT transmitter selects output groups from the network transmission queue, which is ordered FIFO (first-in-first-out) within priority. For data sets created on the local node, the network SYSOUT transmitter creates network data set headers. The network SYSOUT transmitter transmits the data set headers, along with a job header, job trailer, and data set records over an NJE line to the destination node specified in the routing information for the data set. This transmission can be either store-and-forward through intermediate nodes, or direct transmissions to the destination node.

JES2 attempts to transmit as one job, any non-spin SYSOUT which are created under a single job and directed to a common destination (or along a common path). This processing is performed even if multiple SYSOUT data sets are grouped into different output groups with different output priorities.

At the other node, a network SYSOUT receiver receives the NJE headers and data sets. The network SYSOUT receiver uses the information in the NJE job header to create a new job on the receiving system. As a result of this processing, the network SYSOUT receiver places a job on the output queue.The output processor on the receiving node treats this job as a locally completed job. The output processor creates output groups describing the data set, thus making it ready for local print/punch processing or further network transmission, depending on the routing specified for the data sets.