Data field in userData, autoscaling

Hi, guys!
need tour help with setting up the druid autoscale
for now I have this autoscale config:
{
“selectStrategy”: {},
“autoScaler”: {
“type”: “ec2”,
“minNumWorkers”: 1,
“maxNumWorkers”: 2,
“envConfig”: {
“availabilityZone”: “us-east-1d”,
“nodeData”: {
“amiId”: “ami-xxxx”,
“instanceType”: “i3.4xlarge”,
“minInstances”: 1,
“maxInstances”: 2,
“securityGroupIds”: [“sg-xxxx”, “sg-yyyyy”],
“keyName”: “blablalba-key-name”
},
“userData”: {
“impl”: “galaxy”,
“data”: “<SOME_SCRIPT>”,
“env”: null,
“version”: “1”,
“type”: null
}
}
}
}
I have problems with ‘data’ field in ‘userData’ object
why do we need it?
as I understand, we can write some shell command there, and it will be executed after the instance launched, right? or no?
for example, when I send config to the server with such data…
{

“data”: “sudo touch /opt/example.txt”

}
… I expect to find the example.txt file in the opt dir, but there is no file there =( that is, it doesn’t work?

even more
the overlord process doesn’t show me the data field in the logs:
http://joxi.ru/Q2KaNXvILYoMEr
and when I try to see this config in the console – there are only impl, type, version and env fields, that’s all, no data field

where could the problem be? Or didn’t I understand the true meaning of this field?

I just need to connect the new nodes (which were created during the scaling) to the druid cluster, how can I do this without this field?