Painless new behaviour for nulls in version 7.x

Elastic Stack Elasticsearch
1

Hi all,

After the upgrade from version 6 to 7, we are experiencing a new behavior when processing aggregation. Now, whenever the Elastic comes across with a null value in a field while running a Painless script it breaks, by default.

For example, the following simple if/else statement:

if(doc['MY_FIELD'].value <= 9000) 
{return 'Less than 9k' }
else if(doc['MY_FIELD'].value > 9000)
{ return ''More than 9k' }

Until version 6, we could run this code without any issue whatsoever. However now, if you try the same thing it fails when Elastic processes a single null value for the field.

  "lang" : "painless",
  "caused_by" : {
    "type" : "illegal_state_exception",
    "reason" : "A document doesn't have a value for a field! Use doc[<field>].size()==0 to check if a document is missing a field!"

That makes now mandatory to stuff this "doc[<field>].size()==0" checking everywhere.

Is that a parameter to shut this new behavior off and resume the previous one?
Thanks

2

This behavior is expected starting in 7.0 (see https://github.com/elastic/elasticsearch/issues/29286). Using the size() == 0 check for missing values is intentional. While in the past a "default" value would have been provided, this results in trappiness for users not knowing that some documents may not have a value. The error message you received is to ensure users do not fall into that trap.