Help Center/ ModelArts/ FAQs/ Dedicated Resource Pool/ What Should I Do When Model Deployment Stuck in 'Initializing Environment' Due to Kubeinfer Plugin Errors
Updated on 2026-07-02 GMT+08:00

What Should I Do When Model Deployment Stuck in 'Initializing Environment' Due to Kubeinfer Plugin Errors

Model deployment in the ModelArts dedicated resource pool requires the KubeInfer plugin to be in the running state. If the plugin is abnormal, the model deployment job will remain stuck in the 'Initializing environment' state. If the KubeInfer plugin is abnormal, perform the following steps:

  1. Verify that the resource pool meets the plugin's quota requirements. You can check the resource usage of all nodes on the Nodes tab of the resource pool details page.
  2. Ensure the number of nodes in the resource pool is not less than the number of plug-in replicas. If it is, edit the plugin and set the replica count to match the current number of nodes to restore services.