DSXchange

Is their any possible way to convert an Datastage Server Job to parallel extender, (ie., to implement the parallelism concept in server jobs).

Pipeline parallelism is implemented using inter-process row buffering.
Partition parallelism is implemented using either Transformer or Link Partitioner stage to partition the data and Link Collector stage should you need to collect the parallel streams back to a single stream.

In short, in parallel jobs parallelism is automatic and scales without re-design or re-compilation. In server jobs parallelism must be designed in, and therefore can not be scaled differently without re-design.