Known solutions and recommendations can help you when your update process to scale a deployment fails.

Problem

An update process to scale deployments fails.

Cause

You might be attempting to initiate an update process to scale a deployment that previously failed to scale.

Procedure

  1. The cloud administrator must delete the new virtual machines for all the clustered nodes of the deployment.
  2. The cloud administrator must delete the virtual machines that were not removed during the scale in update process.
  3. Initiate another scale out update process on the previously failed deployment.

    Application Services prepopulates the previously applied successful property values. The new values are taken from the previous failed deployment.

    For example, if you initiate an update process to scale an AppServer1 clustered node from 1 to 2 nodes and scale an AppServer2 clustered node from 1 to 3 nodes and the update process failed for AppServer2 node. In the second update process, the value for AppServer1 is populated to 2 and for AppServer2 node you can specify a value greater than the previous value.

  4. Initiate another scale in update process if the deployment failed and the virtual machines were removed.
  5. Initiate another update process to scale out all the failed nodes, so that the update scripts run on all of the dependent components such as load balancer, to make the components valid.