This issue has been resolved. The VM at 34.23.12.208 became temporarily unresponsive due to a network stack issue at the VM level. There were no host-level GCP incidents, reboots, or service crashes during the period. External monitoring reported “No packets returned by host,” and connectivity was restored after rebooting the VM.