I am trying to deploy ML changes via gradle from a central management cloud pc.

I could successful do a fresh deploy on a new ML installation on that cloud pc. However I could not deploy to an ML server on Azure due to connection timeout. (I have try to run the same gradlw -i mlDeploy more than 10 times. They all result the same error.)

I notice most of the other things like content db, module db, schema db, app servers, users, roles are all successfully deployed. I suspect it might to to the mlLoadModules.

If I run gradlew -i mlLoadModules to that remote Azure ML server. I will get the same time out error.

How to resolve the problem? Can I change the timeout setting in mlgradle? (I really do not want to run the Gradle from that Azure ML server. I prefer that central deployment node concept.)

CodePudding user response：

The issue has been resolved. It is not a mlgradle problem.

The issue was caused by the firewall on the Azure ML host. It seems mlLoadModules needs to communicate with the application REST port to upload the Xquery programs. That application REST port is not yet configured to be open in the default Azure ML ARM template from the marklogic official image.

After openning the application REST point, for my case, it is PORT 8030, I could run gradlew -i mlDeploy successfully.