I've got error spinup.sh.

[jerryjung]

I try to setting Sparkline BI Accelerator cluster on aws for test purpose.
I used cluster spinup tool.
I've got error below and further processing is unable.
It seems to aws error, how can I fix it?

Config Files added to psb.
Waiting for all nodes to fully spin up (this might take a while)...
Cluster fully spun up!
Uploading provision script to s3://spark-druid-spinup/psf.sh...
upload: ../../../../../../tmp/tmp_provision.sh to s3://spark-druid-spinup/psf.sh
Running provision script on all instances...
Running script on 3 EC2 instance(s)...
Received: --document-name AWS-RunShellScript --timeout-seconds 600 --instance-ids i-00f61d44c7e90a012 i-07949721fa02ae984 i-08373165a228c4df0 --parameters file:///tmp/params.json --comment PSF

An error occurred (InvalidInstanceId) when calling the SendCommand operation:
Error(s) occurred; exiting.

I also checked ssm infolist is empty.

aws ssm describe-instance-information
{
    "InstanceInformationList": []
}

[hbutani]

Sorry for the delay. The person who did this work is very busy with end-of-term exams right now, so please forgive us for the delays.

Here are some suggestions from him:

• one or more EC2 instances die or fail to spin up because of low spot prices (pretty unlikely but possible)
• the ssm agent isn't installing properly because something changed in AWS
• something about the security group/IAM policies not being configured properly (I don't think this is it)
• or, something else in AWS changed and broke the script

If you have already checked all the above; can you post the configuration file you used.

[jerryjung]

Thank you for your response, I checked already ssm agent installation and security group.
There are no any problems. I assumed that ssm communication makes the problem.
I'll attach my configuration files.

configFile-TPCHSample.txt
spinup.sh.txt

Anyway, I provisioned manually and testing Spakline cluster. There is no error.