def ha_appliances_with_providers(ha_multiple_preupdate_appliances, app_creds):
_, _, appl2 = configure_appliances_ha(ha_multiple_preupdate_appliances,
app_creds["password"])
# Add infra/cloud providers and create db backup
provider_app_crud(VMwareProvider, appl2).setup()
provider_app_crud(OpenStackProvider, appl2).setup()
return ha_multiple_preupdate_appliances
def test_appliance_console_restore_db_ha(request, unconfigured_appliances,
app_creds):
"""Configure HA environment with providers, run backup/restore on configuration,
Confirm that ha failover continues to work correctly and providers still exist.
Polarion:
assignee: jhenner
caseimportance: high
casecomponent: Appliance
initialEstimate: 1/4h
Bugzilla:
1740515
"""
pwd = app_creds["password"]
appl1, appl2, appl3 = configure_appliances_ha(unconfigured_appliances, pwd)
# Add infra/cloud providers and create db backup
provider_app_crud(VMwareProvider, appl3).setup()
provider_app_crud(EC2Provider, appl3).setup()
appl1.db.backup()
providers_before_restore = set(appl3.managed_provider_names)
# Restore DB on the second appliance
appl3.evmserverd.stop()
appl1.rh_postgresql95_repmgr.stop()
appl2.rh_postgresql95_repmgr.stop()
appl1.db.drop()
appl1.db.create()
fetch_v2key(appl3, appl1)
restore_db(appl1)
appl1.appliance_console.reconfigure_primary_replication_node(pwd)
appl2.appliance_console.reconfigure_standby_replication_node(
pwd, appl1.hostname)
appl3.appliance_console.configure_automatic_failover(
primary_ip=appl1.hostname)
appl3.evm_failover_monitor.restart()
appl3.evmserverd.start()
appl3.wait_for_web_ui()
# Assert providers still exist after restore
assert providers_before_restore == set(appl3.managed_provider_names), (
'Restored DB is missing some providers')
with LogValidator(evm_log,
matched_patterns=['Starting to execute failover'],
hostname=appl3.hostname).waiting(timeout=450):
# Cause failover to occur
appl1.db_service.stop()
appl3.evmserverd.wait_for_running()
appl3.wait_for_web_ui()
# Assert providers still exist after ha failover
assert providers_before_restore == set(appl3.managed_provider_names), (
'Restored DB is missing some providers')
# Verify that existing provider can detect new VMs after restore/failover
virtual_crud = provider_app_crud(VMwareProvider, appl3)
vm = provision_vm(request, virtual_crud)
assert vm.mgmt.is_running, "vm not running"
def test_appliance_console_restore_db_ha(request, unconfigured_appliances,
app_creds):
"""Configure HA environment with providers, run backup/restore on configuration,
Confirm that ha failover continues to work correctly and providers still exist.
Polarion:
assignee: jhenner
caseimportance: high
casecomponent: Appliance
initialEstimate: 1/4h
Bugzilla:
1693189
1740515
1856470
"""
pwd = app_creds["password"]
db_appl_1, db_appl_2, webui_appl = configure_appliances_ha(
unconfigured_appliances, pwd)
# Add infra/cloud providers and create db backup
provider_app_crud(VMwareProvider, webui_appl).setup()
provider_app_crud(OpenStackProvider, webui_appl).setup()
db_appl_1.db.backup()
providers_before_restore = set(webui_appl.managed_provider_names)
# Restore DB on the second appliance
webui_appl.evmserverd.stop()
db_appl_1.repmgr.stop()
db_appl_2.repmgr.stop()
db_appl_1.db.drop()
db_appl_1.db.create()
fetch_v2key(webui_appl, db_appl_1)
restore_db(db_appl_1)
db_appl_1.appliance_console.reconfigure_primary_replication_node(pwd)
db_appl_2.appliance_console.reconfigure_standby_replication_node(
pwd, db_appl_1.hostname)
# Reboot the standby node to make sure the repmgr10 starts automatically and is ready to
# take-over (BZ 1856470).
db_appl_2.reboot(wait_for_miq_ready=False)
webui_appl.appliance_console.configure_automatic_failover(
primary_ip=db_appl_1.hostname)
webui_appl.evm_failover_monitor.restart()
webui_appl.evmserverd.start()
webui_appl.wait_for_miq_ready()
# Assert providers still exist after restore
assert providers_before_restore == set(
webui_appl.managed_provider_names), (
'Restored DB is missing some providers')
with LogValidator(evm_log,
matched_patterns=['Starting to execute failover'],
hostname=webui_appl.hostname).waiting(timeout=450):
# Cause failover to occur. Note that reboot of the active node may not cause failover to
# occur as the node can come back again soon enough.
db_appl_1.db_service.stop()
webui_appl.evmserverd.wait_for_running()
webui_appl.wait_for_miq_ready()
# Assert providers still exist after ha failover
assert providers_before_restore == set(
webui_appl.managed_provider_names), (
'Restored DB is missing some providers')
# Verify that existing provider can detect new VMs after restore/failover
virtual_crud = provider_app_crud(VMwareProvider, webui_appl)
vm = provision_vm(request, virtual_crud)
assert vm.mgmt.is_running, "vm not running"