Hi. Wonder if someone could help. I have a tMap(tMap1) where the main output goes to tMSSQLOutput. However now I am handling the rejected rows from tMap. So the rejected output from tMap1 goes to tMap2, tMap2 also has an input lookup. The output of tMap1 and tMap2 have same schema. I want to merger the output of tMap1 and tMap2 together (union) and send one union output to tMSSQLOutput. I tried tUnite, but tUnite only accepts one input. tUnite basically works when you are iterating (say a folder full of files) but my requirement here is to perform the union. Any ideas? Thanks, Prasad
Output to the tHashOutput components from your tMaps and then in another subjob use the tUnite to merge the datasets together.
Once you have separated a single flow into 2 or more flows, you cannot re-merge them in the same subjob. So to get round this, you store the data in memory (for relatively small datasets) and merge them together in another subjob.
Thanks. So the output of tMap1 goes to tHashOutput1 and output of tmap2 goes to tHasmOutput2? And in sub job I have two tHashInput components linked to tHashoutput1 and 2 respectively? Not entirely clear on how the subjob should be configured or do I have two sub jobs one for each Hashinput?
You would have 2 subjobs; 1 to load the data into the tHashOutput components and 1 to merge the data. You can use a tUnite for this so long as the schemas of both the datasets are the same.
This was the right answer, but I have to mention that in 7.1 and 7.2 versions we have to enable the hash elements under: Project Settings --> Designer --> Pallete Settings --> Technical
