I am working on a Node batch process that uses a N1QL query to stream row data to a SQL database and a CSV (for back-up purposes). According to the Node 2.6 SDK Documentation, “If you are expecting a large result set from the server you can omit passing the callback to the query() function and instead register to the returned object’s row , end , and error events.”
However, based on my local testing, if an error event is received no further rows are processed. What is the best to either restart or continue processing row information? One other note: the current code is using the N1QL example shown in the documentation:
This feature will be available in NodeJS SDK 3.2 which will be the next minor release of our SDK. Within a month. Will work exactly the same as SDK 2x
I am seeing what I can do in order to get this feature documented as it is in our SDK 2x docs. You can use the new SDK 3.2 but you would have to build the binaries locally until we release version 3.2 I can get you instructions on how to do that.
Can you let me know what OS you use on your dev machine, this will assist me in providing you the right instructions unless you would like to wait for the SDK 3.2 release instead.