def test_heat_vm_scale_after_actions(vm_scaling_stack, actions):
"""
Test VM auto scaling with swact:
Create heat stack for auto scaling using NestedAutoScale.yaml, swact and perform vm scale up and down.
Test Steps:
- Create a heat stack for auto scaling vm ()
- Verify heat stack is created successfully
- Verify heat resources are created
- live migrate the vm if not sx
- cold migrate the vm if not sx
- swact if not sx
- reboot -f vm host
- trigger auto scale by boosting cpu usage in the vm (using dd)
- verify it scale up to the max number of vms (3)
- trigger scale down by killing dd in the vm
- verify the vm scale down to min number (1)
- Delete Heat stack and verify resource deletion
"""
stack_name, vm_id = vm_scaling_stack
vm_helper.wait_for_vm_pingable_from_natbox(vm_id=vm_id)
if not system_helper.is_aio_simplex():
actions = actions.split('-')
if "swact" in actions:
LOG.tc_step("Swact before scale in/out")
host_helper.swact_host()
if "live_migrate" in actions:
LOG.tc_step("live migrate vm before scale in/out")
vm_helper.live_migrate_vm(vm_id)
if "cold_migrate" in actions:
LOG.tc_step("cold migrate vm before scale in/out")
vm_helper.cold_migrate_vm(vm_id)
if "host_reboot" in actions:
if system_helper.is_aio_simplex():
host_helper.reboot_hosts('controller-0')
vm_helper.wait_for_vm_status(vm_id,
status=VMStatus.ACTIVE,
timeout=600,
check_interval=10,
fail_ok=False)
vm_helper.wait_for_vm_pingable_from_natbox(
vm_id, timeout=VMTimeout.DHCP_RETRY)
else:
LOG.tc_step("evacuate vm before scale in/out")
vm_host = vm_helper.get_vm_host(vm_id=vm_id)
vm_helper.evacuate_vms(host=vm_host, vms_to_check=vm_id)
LOG.tc_step(
"Wait for {} vms to auto scale out to {} after running dd in vm(s)".
format(stack_name, 3))
vm_helper.wait_for_auto_vm_scale_out(stack_name, expt_max=3)
LOG.tc_step(
"Wait for {} vms to auto scale in to {} after killing dd processes in vms"
.format(stack_name, 1))
vm_helper.wait_for_auto_vm_scale_in(stack_name, expt_min=1)
def test_system_persist_over_host_reboot(host_type, stx_openstack_required):
"""
Validate Inventory summary over reboot of one of the controller see if data persists over reboot
Test Steps:
- capture Inventory summary for list of hosts on system service-list and neutron agent-list
- reboot the current Controller-Active
- Wait for reboot to complete
- Validate key items from inventory persist over reboot
"""
if host_type == 'controller':
host = system_helper.get_active_controller_name()
elif host_type == 'compute':
if system_helper.is_aio_system():
skip("No compute host for AIO system")
host = None
else:
hosts = system_helper.get_hosts(personality='storage')
if not hosts:
skip(msg="Lab has no storage nodes. Skip rebooting storage node.")
host = hosts[0]
LOG.tc_step("Pre-check for system status")
system_helper.wait_for_services_enable()
up_hypervisors = host_helper.get_up_hypervisors()
network_helper.wait_for_agents_healthy(hosts=up_hypervisors)
LOG.tc_step("Launch a vm")
vm_id = vm_helper.boot_vm(cleanup='function')[1]
vm_helper.wait_for_vm_pingable_from_natbox(vm_id)
if host is None:
host = vm_helper.get_vm_host(vm_id)
LOG.tc_step("Reboot a {} node and wait for reboot completes: {}".format(host_type, host))
HostsToRecover.add(host)
host_helper.reboot_hosts(host)
host_helper.wait_for_hosts_ready(host)
LOG.tc_step("Check vm is still active and pingable after {} reboot".format(host))
vm_helper.wait_for_vm_status(vm_id, status=VMStatus.ACTIVE, fail_ok=False)
vm_helper.wait_for_vm_pingable_from_natbox(vm_id=vm_id, timeout=VMTimeout.DHCP_RETRY)
LOG.tc_step("Check neutron agents and system services are in good state after {} reboot".format(host))
network_helper.wait_for_agents_healthy(up_hypervisors)
system_helper.wait_for_services_enable()
if host in up_hypervisors:
LOG.tc_step("Check {} can still host vm after reboot".format(host))
if not vm_helper.get_vm_host(vm_id) == host:
time.sleep(30)
vm_helper.live_migrate_vm(vm_id, destination_host=host)
def test_reboot_standby_controller(no_simplex):
active, standby = system_helper.get_active_standby_controllers()
LOG.tc_step("'sudo reboot -f' from {}".format(standby))
host_helper.reboot_hosts(standby,
wait_for_offline=True,
wait_for_reboot_finish=True,
force_reboot=True)
system_helper.wait_for_hosts_states(standby,
timeout=360,
check_interval=30,
availability=['available'])
kube_helper.wait_for_pods_healthy(check_interval=30, all_namespaces=True)
def sys_uncontrolled_swact(number_of_times=1):
"""
This is to identify the storage nodes and turn them off and on via vlm
:return:
"""
for i in range(0, number_of_times):
active, standby = system_helper.get_active_standby_controllers()
LOG.tc_step("Doing iteration of {} of total iteration {}".format(
i, number_of_times))
LOG.tc_step("'sudo reboot -f' from {}".format(standby))
host_helper.reboot_hosts(hostnames=active)
LOG.tc_step("Check vms status after controller swact")
vms = get_all_vms()
vm_helper.wait_for_vms_values(vms, fail_ok=False, timeout=600)
for vm in vms:
vm_helper.wait_for_vm_pingable_from_natbox(vm_id=vm)
def sys_reboot_standby(number_of_times=1):
"""
This is to identify the storage nodes and turn them off and on via vlm
:return:
"""
timeout = VMTimeout.DHCP_RETRY if system_helper.is_aio_system(
) else VMTimeout.PING_VM
for i in range(0, number_of_times):
active, standby = system_helper.get_active_standby_controllers()
LOG.tc_step("Doing iteration of {} of total iteration {}".format(
i, number_of_times))
LOG.tc_step("'sudo reboot -f' from {}".format(standby))
host_helper.reboot_hosts(hostnames=standby)
LOG.tc_step("Check vms status after stanby controller reboot")
vms = get_all_vms()
vm_helper.wait_for_vms_values(vms, fail_ok=False, timeout=600)
for vm in vms:
vm_helper.wait_for_vm_pingable_from_natbox(vm_id=vm,
timeout=timeout)
def _test_status_firewall_reboot():
"""
Test iptables status after reboot of controller
Test Steps:
- Stop iptables service
- Confirm iptables service has stopped
- Reboot the controller being tested
- Confirm iptables service is online
- Repeat for second controller
"""
LOG.tc_step("Getting the controller(s)")
controllers = system_helper.get_controllers()
for controller in controllers:
with host_helper.ssh_to_host(controller) as con_ssh:
LOG.tc_step("Stopping iptables service")
cmd = 'service iptables stop'
con_ssh.exec_sudo_cmd(cmd)
LOG.tc_step("checking iptables status")
cmd = 'service iptables status'
code, output = con_ssh.exec_sudo_cmd(cmd)
assert 'Active: inactive' or 'Active: failed' in output, "iptables service did not stop running on host {}"\
.format(controller)
LOG.tc_step("Rebooting {}".format(controller))
HostsToRecover.add(controller)
host_helper.reboot_hosts(controller)
with host_helper.ssh_to_host(controller) as con_ssh:
LOG.tc_step(
"Checking iptables status on host {} after reboot".format(
controller))
cmd = 'service iptables status | grep --color=never Active'
code, output = con_ssh.exec_sudo_cmd(cmd)
assert 'active' in output, "iptables service did not start after reboot on host {}".format(
controller)
def reboot_hosting_node(vm_type, vm_id, force_reboot=False):
host = vm_helper.get_vm_host(vm_id)
host_helper.reboot_hosts(host, force_reboot=force_reboot)
rescue_vm(vm_type, vm_id)
