hdc create-cluster fails with blueprint not found error in the first attempt

[#1]

I brought up a cloud controller using AWS CLI. After ssh-ing to the controller instance I have downloaded hdc cli jar.

Configured hdc with server address, username and password. Now the attempt to create a cluster fails with Blueprint not found error. If I just login to the HDC UI once and then retry the same command via hdc cli it goes through fine too. I checked ~/.hdc/config file server, user and password are all correct. Has anyone faced similar issue or any clue what I may be doing wrong. Which log files can i check to get more info on what is happening when we issue a command on cli?

./hdc create-cluster -cli-input-json /tmp/hdc-cli-deploy.json --wait true
ERROR: status code: 404, message: Blueprint 'EDW-ETL: Apache Hive 1.2.1, Apache Spark 2.1' not found.  

[#2 · Accepted solution]

Hi @dbalasundaran,

Default resources like Blueprints are created at the first GET request to the desired resource endpoint. When you used the UI, the default resources were created that time. After that the hdc create-cluster probably worked.

You always needs to trigger the creation of default blueprints if you use them. One option is to use the UI, other option is to run "hdc list-cluster-types" that will create the default blueprints for you.

[#3]

Yes, hdc list-cluster-types works @Tamas Bihari. Thankyou!