Have the ability for an integration to skip mapping which are null
L
Linas Daneliukas
In some cases we have 2 exporters/intergation that are enriching the same entity.
In some case, one of the exporter will set a value and then the second exporter will set the value as null.
This is causing the value to flip between the value and null.
I want to be able to add a flag in the mapping
ignoreNulls: true/false
When this is set to true, the exporter/integration will not update the field in the entity.
This could be applied to Ocean based integration as it should upsert only the mappings with has a value.
L
Linas Daneliukas
This would allow clients to have a simplistic method of weighted datasources. A simple use case would be this:
Datasource "Puppet Agent" is the best source for server OS names as it gives me "Rocky Linux" - however puppet agent is not always running on all machines.
Datasource "OpenStack" is also a decent source for server OS names but it only knows them for some servers, not all of them.
Server A is hosted on AWS and has Puppet Agent - I see "Rocky Linux" in Port.
Server B is on OpenStack (which in this case, does not know its OS so it gives out null) and has Puppet Agent - in Port I see "Rocky Linux", then I see "null", then "Rocky Linux", then "null" as both datasources fight each other.
It'd be great to be able to tell the OpenStack exporter to ignore values that are
null
instead of shipping a literal null
since these won't add to the data I might have from other (better) sources, but it will make the data wrong.The mapping tester already removes a property if the value of it ends up a
null
but thats only the tester being misleading - the real ingestions does add the nulls.