utils/publish.py search_ingest should log http.response from the search ingest lambda

[ericblau]

Evidently it is possible to get a 200 response from Globus Search ingest, with zero documents ingested. (This can happen if the index has exceeded its 1000 fields--but the error message will only be in the Globus Search Task.) Thus, the dlhub_sdk search_ingest function should log the response from Globus Search ingest even when "successful" so that the user has the Globus Search taskid to try to figure out why their document doesn't show up in the index.