Symptoms
VM location and resource usage is not updated in Cloud Infrastructure after VM migration to another Virtualization Node. The following messages related to migrated VM are observed in /var/log/paciagent.log on the destination Virtualizaiton Hardware Node:
2019-03-08 01:38:12,558 DBG srv_handler: got event PET_DSP_EVT_VM_ADDED|PET_DSP_EVT_VM_CREATED from \{%VM_UUID%} @[agent.c][847][srv_handler][39885])
2019-03-08 01:38:12,752 ERR PrlVm_SubscribeToPerfStats: failure PRL_ERR_VM_CONFIG_INVALID_SERVER_UUID @[agent.c][1256][create_ve_usage][39885])
...
2019-03-08 01:39:11,326 INF PrlVm_SubscribeToPerfStats: retrying ... @[agent.c][1262][create_ve_usage][39885])
2019-03-08 01:39:13,331 ERR PrlVm_SubscribeToPerfStats: failure PRL_ERR_VM_CONFIG_INVALID_SERVER_UUID @[agent.c][1256][create_ve_usage][39885])
2019-03-08 01:39:13,335 ERR create_ve_usage: failure PRL_ERR_VM_CONFIG_INVALID_SERVER_UUID @[agent.c][875][srv_handler][39885])
Cause
ServerUUID of hardware node is not updated yet in configuration file of migrated VM on Virtuozzo Hardware Node.
Resolution
Handling of such behavior on Cloud Infrastructure side is improved in scope of the request CCU-17070 since Cloud Infrastructure 17.12. Contact CloudBlue technical support in case if behavior is observed with the latest releases of Cloud Infrastructure.
Perform snmpd service restart on the destination Virtualization Node to re-gather performance statistics and update location for migrated VM in Cloud Infrastructure.