/

Troubleshooting Autoscale Deployment - General


General autoscale setup issues and solutions:

Stream Manager Red5 Pro Service Failure

If the red5pro jsvc service is failing to start on your stream manager, verify that your LICENSE.KEY file is in the red5pro root directory, and that the stream manager has access to the database.

Multiple Autoscale Environments

VERY IMPORTANT: If you have more than one autoscaling environment running, it is essential that the instancecontroller.instanceNamePrefix= property is set to a completely unique value for each environment, or the first stream manager may terminate the nodes belonging to the second group.

As an example, if you have one environment with instancecontroller.instanceNamePrefix=node and another one with instancecontroller.instanceNamePrefix=node123, the node123 instances will be terminated by the first stream manager. This occurs because it identifies them as starting with node and therefore as instances it should be managingm which are not present it its database. Find the instancecontroller.instanceNamePrefix= in the {red5pro}/webapps/streammanager/WEB-INF/red5-web.properties, and edit it accordingly.

Instance Quotas

Most cloud platforms have regional and/or global quotas for VMs (especially with new accounts). Make sure that your account has sufficient allowance in the region(s) where you plan to depoloy your autoscaling solution to allow for growth.

Nodgroup Doesn't Populate

If you have created a new nodegroup, but adding an Origin server doesn't automatically spin up an accompanying Edge server, check the following on your node image:

  • Verify that red5pro/conf/autoscale.xml in your Red5Pro server image points to the IP address or URL of your stream manager

    replace 0.0.0.0 in <property name="cloudWatchHost" value="http://0.0.0.0:5080/streammanager/cloudwatch"/>

    And make sure that you have edited the active value from default “false” to “true” (<property name="active" value="true"/>)

  • Check {red5pro}/conf/cluster.xml - make sure that the password set for <!-- edge/origin link cluster password --> matches the cluster.password= in the stream manager red5-web.properties file.
  • Make sure that your have your LICENSE.KEY file in the {red5pro} root directory; clustering (and in turn, autoscaling) will not work without a valid license key.

If all of the above check out OK, then:

  • On the Stream Manager: modify conf/logback.xml, setting <logger name="com.red5pro.services.streammanager" level="DEBUG" />; restart the service on the stream manager to apply the new log level and tail {red5pro}/log/red5.log file.
  • On the node VM, modify conf/logback.xml, setting <logger name="com.red5pro.clustering.autoscale" level="DEBUG"/> then start red5pro on your instance by changing to the red5pro home directory (default: /usr/local/red5pro) and running sudo ./red5.sh

You should see log messages about an instance pinging the Stream Manager and requesting a role assignment. If you don't, then check for inbound port access to your stream manager (or outbound port access from your node security group).