COA Troubleshooting & Problem Resolution 2 — Questions and Answers
Question 1: An instance is stuck in 'ERROR' state after a failed live migration. Which command retrieves the error details from the Nova compute log?
- nova show <instance-id>
- openstack server show <instance-id> (Correct answer)
- nova-manage vm list
- openstack compute service list
Correct answer: openstack server show <instance-id>
The 'openstack server show' command displays fault messages and the last error that caused an instance to enter ERROR state.
Question 2: After updating the Neutron ML2 plugin configuration, the neutron-server fails to start. What is the first log file you should inspect?
- /var/log/neutron/openvswitch-agent.log
- /var/log/neutron/server.log (Correct answer)
- /var/log/neutron/dhcp-agent.log
- /var/log/neutron/l3-agent.log
Correct answer: /var/log/neutron/server.log
The neutron-server.log captures startup errors including plugin misconfiguration, making it the primary source for server-start failures.
Question 3: A tenant reports that their floating IP is assigned but they cannot reach the instance from the internet. Which Neutron resource should you verify first?
- Security group egress rules
- Router external gateway association (Correct answer)
- DHCP agent binding
- Network QoS policy
Correct answer: Router external gateway association
A floating IP requires a router with an external gateway set; without it, NAT cannot route traffic between the external network and the tenant network.
Question 4: You run 'openstack volume list' and see a volume stuck in 'attaching' state for over 30 minutes. What is the recommended corrective action?
- Delete and recreate the volume
- Run openstack volume set --state available <vol-id> (Correct answer)
- Restart the nova-conductor service
- Detach the volume with --force flag in Nova
Correct answer: Run openstack volume set --state available <vol-id>
Forcing the Cinder volume state back to 'available' with 'openstack volume set --state' unblocks stuck attach operations without data loss.
Question 5: Which command verifies that all Nova compute services are up and reporting correctly across the deployment?
- nova host-list
- openstack compute service list (Correct answer)
- nova-manage service list
- openstack hypervisor list
Correct answer: openstack compute service list
'openstack compute service list' shows the state and status of every Nova service node, including whether each is 'up' or 'down'.
Question 6: A Keystone token validation is failing with 'Service Unavailable'. The Keystone service is running but errors appear in the log referencing the Fernet key repository. What is the likely cause?
- The token provider is set to UUID instead of Fernet
- The Fernet key rotation has invalidated all existing tokens (Correct answer)
- The identity backend is misconfigured
- The PKI certificate has expired
Correct answer: The Fernet key rotation has invalidated all existing tokens
Fernet key rotation invalidates tokens signed with older keys; if rotation occurs too aggressively before tokens expire, active sessions fail validation.
Question 7: An operator finds that Nova is scheduling all new instances to a single compute node despite others being available. Which service configuration option controls the scheduler's host selection algorithm?
- compute_driver in nova.conf
- scheduler_driver in nova.conf
- scheduler_default_filters in nova.conf (Correct answer)
- ram_allocation_ratio in nova.conf
Correct answer: scheduler_default_filters in nova.conf
The 'scheduler_default_filters' option defines which filters (e.g., RamFilter, CoreFilter, ComputeFilter) the Nova scheduler applies when selecting a host.
COA Troubleshooting & Problem Resolution 2
An instance is stuck in 'ERROR' state after a failed live migration.
Which command retrieves the error details from the Nova compute log?