[mattdahse]

When returning an object reference to a search or a search object itself, the getInputData phase of a map/reduce script is not executing the search if there are no defined columns. Adding a column (such as internalid) will correct this, but without that, the map and reduce phases are not called, even when the only information that they need might be the result id.

I can’t find any documentation that would lead me to believe this is intentional. Does anyone know if this is a known defect?
This is a cached copy. Click here to see the original post.

[david.smith]

I found this to be true as well. I just add the internalid and accept it. I have not seen (or looked) for documentation on this.

[chanarbon]

Generally, in map where the values comes from a search, the internalid is tracked as the key of the mapContext so it would be a lot easier to deal with the map execution with the internal id directly accessible on the mapContext object.

[mattdahse]

This is interesting. Now I am curious how things are handled if I include a summary column and the result.id is not available.

---

david.smith replied on 06/02/2017, 12:06 PM: There is no id if you’re using summary. Your results are still passed to the Map stage. Not sure how that would be handled by the Reduce.

[Author]

Posts