Increase Timeout (#377)

- Duruing partner cloud upgrade it was detected that COU was failing to
check if the whole model is active idle because the default time is not
enough.

- Keystone also needs some extra time, and this change from 30 to 40
minutes.

README.md

@@ -44,7 +44,7 @@ Commands:
 - `COU_MODEL_RETRIES` - define how many times to retry the connection to Juju model before giving up. Default value is 5 times.
 - `COU_MODEL_RETRY_BACKOFF` - define number of seconds to increase the wait between connection to the Juju model retry attempts. Default value is 2 seconds.
 - `COU_STANDARD_IDLE_TIMEOUT` - how long COU will wait for an application to settle to active/idle and declare the upgrade complete. The default value is 300 seconds.
-- `COU_LONG_IDLE_TIMEOUT` - a longer version of COU_STANDARD_IDLE_TIMEOUT for applications that are known to need more time than usual to upgrade like such as Keystone and Octavia. The default value is 1800 seconds.
+- `COU_LONG_IDLE_TIMEOUT` - a longer version of COU_STANDARD_IDLE_TIMEOUT for applications that are known to need more time than usual to upgrade like such as Keystone and Octavia. The default value is 2400 seconds.
 
 ## Supported Upgrade Paths
 

cou/apps/base.py

@@ -48,7 +48,7 @@
 STANDARD_IDLE_TIMEOUT: int = int(
     os.environ.get("COU_STANDARD_IDLE_TIMEOUT", 5 * 60)
 )  # default of 5 min
-LONG_IDLE_TIMEOUT: int = int(os.environ.get("COU_LONG_IDLE_TIMEOUT", 30 * 60))  # default of 30 min
+LONG_IDLE_TIMEOUT: int = int(os.environ.get("COU_LONG_IDLE_TIMEOUT", 40 * 60))  # default of 40 min
 ORIGIN_SETTINGS = ("openstack-origin", "source")
 REQUIRED_SETTINGS = ("enable-auto-restarts", "action-managed-upgrade", *ORIGIN_SETTINGS)
 LATEST_STABLE = "latest/stable"

cou/steps/plan.py

@@ -369,9 +369,9 @@ def _get_pre_upgrade_steps(analysis_result: Analysis, args: CLIargs) -> list[Pre
             coro=analysis_result.model.wait_for_active_idle(
                 # NOTE (rgildein): We need to DEFAULT_TIMEOUT so it's possible to change if
                 # a network is too slow, this could cause an issue.
-                # We are using max function to ensure timeout is always at least 11 (1 second
+                # We are using max function to ensure timeout is always at least 120 (110 seconds
                 # higher than the idle_period to prevent false negative).
-                timeout=max(DEFAULT_TIMEOUT + 1, 11),
+                timeout=max(DEFAULT_TIMEOUT, 120),
                 idle_period=10,
                 raise_on_blocked=True,
             ),

docs/how-to/interruption.rst

@@ -33,7 +33,7 @@ Usage example:
             Upgrade software packages on unit 'keystone/2'
         Upgrade 'keystone' to the new channel: 'victoria/stable'
         Change charm config of 'keystone' 'openstack-origin' to 'cloud:focal-victoria'
-        Wait for up to 1800s for model 'test-model' to reach the idle state
+        Wait for up to 2400s for model 'test-model' to reach the idle state
         Verify that the workload of 'keystone' has been upgraded
 
     Would you like to start the upgrade? Continue (y/N): n

docs/how-to/no-backup.rst

@@ -49,7 +49,7 @@ Upgrade:
             Upgrade software packages on unit 'rabbitmq-server/2'
         Upgrade 'rabbitmq-server' to the new channel: '3.9/stable'
         Change charm config of 'rabbitmq-server' 'source' to 'cloud:focal-victoria'
-        Wait for up to 1800s for model 'test-model' to reach the idle state
+        Wait for up to 2400s for model 'test-model' to reach the idle state
         Verify that the workload of 'rabbitmq-server' has been upgraded
 
     Continue (y/n): y

docs/how-to/plan-upgrade.rst

@@ -37,7 +37,7 @@ Output example
                 Change charm config of 'keystone' 'action-managed-upgrade' to 'False'
                 Upgrade 'keystone' to the new channel: 'victoria/stable'
                 Change charm config of 'keystone' 'openstack-origin' to 'cloud:focal-victoria'
-                Wait for up to 1800s for model 'test_model' to reach the idle state
+                Wait for up to 2400s for model 'test_model' to reach the idle state
                 Verify that the workload of 'keystone' has been upgraded on units: keystone/0
         Control Plane subordinate(s) upgrade plan
             Upgrade plan for 'keystone-ldap' to 'victoria'
@@ -59,7 +59,7 @@ Output example
                         ├── Upgrade the unit: 'nova-compute/0'
                         ├── Resume the unit: 'nova-compute/0'
                 Enable nova-compute scheduler from unit: 'nova-compute/0'
-                Wait for up to 1800s for model 'test_model' to reach the idle state
+                Wait for up to 2400s for model 'test_model' to reach the idle state
                 Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/0
         Remaining Data Plane principal(s) upgrade plan
             Upgrade plan for 'ceph-osd' to 'victoria'

docs/how-to/upgrade-cloud.rst

@@ -130,7 +130,7 @@ Usage example
                 Upgrade software packages on unit 'rabbitmq-server/2'
             Upgrade 'rabbitmq-server' to the new channel: '3.9/stable'
             Change charm config of 'rabbitmq-server' 'source' to 'cloud:focal-victoria'
-            Wait for up to 1800s for model 'test-model' to reach the idle state
+            Wait for up to 2400s for model 'test-model' to reach the idle state
             Verify that the workload of 'rabbitmq-server' has been upgraded
         ...
     Would you like to start the upgrade? Continue (y/N): y
@@ -145,7 +145,7 @@ Usage example
             Upgrade software packages on unit 'rabbitmq-server/2'
         Upgrade 'rabbitmq-server' to the new channel: '3.9/stable'
         Change charm config of 'rabbitmq-server' 'source' to 'cloud:focal-victoria'
-        Wait for up to 1800s for model 'test-model' to reach the idle state
+        Wait for up to 2400s for model 'test-model' to reach the idle state
         Verify that the workload of 'rabbitmq-server' has been upgraded
 
     Continue (y/n): y
@@ -158,7 +158,7 @@ Usage example
                 Upgrade software packages on unit 'keystone/2'
             Upgrade 'keystone' to the new channel: 'victoria/stable'
             Change charm config of 'keystone' 'openstack-origin' to 'cloud:focal-victoria'
-            Wait for up to 1800s for model 'test-model' to reach the idle state
+            Wait for up to 2400s for model 'test-model' to reach the idle state
             Verify that the workload of 'keystone' has been upgraded
 
     Continue (y/n): y

docs/reference/environment-variables.rst

@@ -14,4 +14,4 @@ Environment Variables
   to **active/idle** and declare the upgrade complete. The default value is 300 seconds.
 * **COU_LONG_IDLE_TIMEOUT** - a longer version of **COU_STANDARD_IDLE_TIMEOUT** for applications
   that are known to need more time than usual to upgrade, such as Keystone and Octavia. The
-  default value is 1800 seconds.
+  default value is 2400 seconds.

tests/functional/tests/smoke.py

@@ -153,7 +153,7 @@ def generate_expected_plan(self, backup: bool = True) -> str:
             "\t\t\t\tUpgrade software packages on unit 'mysql-innodb-cluster/2'\n"
             "\t\t\tChange charm config of 'mysql-innodb-cluster' 'source' to "
             "'cloud:focal-victoria'\n"
-            "\t\t\tWait for up to 1800s for app 'mysql-innodb-cluster' to reach the idle state\n"
+            "\t\t\tWait for up to 2400s for app 'mysql-innodb-cluster' to reach the idle state\n"
             "\t\t\tVerify that the workload of 'mysql-innodb-cluster' has been upgraded on units: "
             "mysql-innodb-cluster/0, mysql-innodb-cluster/1, mysql-innodb-cluster/2\n"
         )

tests/mocked_plans/sample_plans/base.yaml

@@ -10,7 +10,7 @@ plan: |
                 Change charm config of 'keystone' 'action-managed-upgrade' to 'False'
                 Upgrade 'keystone' to the new channel: 'victoria/stable'
                 Change charm config of 'keystone' 'openstack-origin' to 'cloud:focal-victoria'
-                Wait for up to 1800s for model 'base' to reach the idle state
+                Wait for up to 2400s for model 'base' to reach the idle state
                 Verify that the workload of 'keystone' has been upgraded on units: keystone/0
         Control Plane subordinate(s) upgrade plan
             Upgrade plan for 'keystone-ldap' to 'victoria'
@@ -32,7 +32,7 @@ plan: |
                         ├── Upgrade the unit: 'nova-compute/0'
                         ├── Resume the unit: 'nova-compute/0'
                 Enable nova-compute scheduler from unit: 'nova-compute/0'
-                Wait for up to 1800s for model 'base' to reach the idle state
+                Wait for up to 2400s for model 'base' to reach the idle state
                 Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/0
         Remaining Data Plane principal(s) upgrade plan
             Upgrade plan for 'ceph-osd' to 'victoria'

tests/unit/apps/test_auxiliary.py

@@ -168,9 +168,9 @@ def test_auxiliary_upgrade_plan_ussuri_to_victoria_change_channel(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -238,9 +238,9 @@ def test_auxiliary_upgrade_plan_ussuri_to_victoria(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -314,9 +314,9 @@ def test_auxiliary_upgrade_plan_ussuri_to_victoria_ch_migration(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -694,9 +694,9 @@ def test_ceph_mon_upgrade_plan_xena_to_yoga(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -772,9 +772,9 @@ def test_ceph_mon_upgrade_plan_ussuri_to_victoria(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -1119,9 +1119,9 @@ def test_mysql_innodb_cluster_upgrade(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for app '{app.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for app '{app.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=[app.name]),
+            coro=model.wait_for_active_idle(2400, apps=[app.name]),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "

tests/unit/apps/test_core.py

@@ -230,9 +230,9 @@ def test_upgrade_plan_ussuri_to_victoria(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -314,9 +314,9 @@ def test_upgrade_plan_ussuri_to_victoria_ch_migration(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -392,9 +392,9 @@ def test_upgrade_plan_channel_on_next_os_release(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -471,9 +471,9 @@ def test_upgrade_plan_origin_already_on_next_openstack_release(model):
             coro=model.upgrade_charm(app.name, "victoria/stable"),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -588,9 +588,9 @@ def test_upgrade_plan_application_already_disable_action_managed(model):
             ),
         ),
         PostUpgradeStep(
-            description=f"Wait for up to 1800s for model '{model.name}' to reach the idle state",
+            description=f"Wait for up to 2400s for model '{model.name}' to reach the idle state",
             parallel=False,
-            coro=model.wait_for_active_idle(1800, apps=None),
+            coro=model.wait_for_active_idle(2400, apps=None),
         ),
         PostUpgradeStep(
             description=f"Verify that the workload of '{app.name}' has been upgraded on units: "
@@ -809,7 +809,7 @@ def test_nova_compute_upgrade_plan(model):
         Enable nova-compute scheduler from unit: 'nova-compute/0'
         Enable nova-compute scheduler from unit: 'nova-compute/1'
         Enable nova-compute scheduler from unit: 'nova-compute/2'
-        Wait for up to 1800s for model 'test_model' to reach the idle state
+        Wait for up to 2400s for model 'test_model' to reach the idle state
         Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/0, nova-compute/1, nova-compute/2
     """  # noqa: E501 line too long
     )
@@ -862,7 +862,7 @@ def test_nova_compute_upgrade_plan_single_unit(model):
                 ├── Upgrade the unit: 'nova-compute/0'
                 ├── Resume the unit: 'nova-compute/0'
         Enable nova-compute scheduler from unit: 'nova-compute/0'
-        Wait for up to 1800s for model 'test_model' to reach the idle state
+        Wait for up to 2400s for model 'test_model' to reach the idle state
         Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/0
     """
     )

tests/unit/steps/test_hypervisor.py

@@ -420,7 +420,7 @@ def test_hypervisor_upgrade_plan(model):
             Wait for up to 300s for app 'cinder' to reach the idle state
             Verify that the workload of 'cinder' has been upgraded on units: cinder/0
             Enable nova-compute scheduler from unit: 'nova-compute/0'
-            Wait for up to 1800s for model 'test_model' to reach the idle state
+            Wait for up to 2400s for model 'test_model' to reach the idle state
             Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/0
         Upgrade plan for 'az-1' to 'victoria'
             Disable nova-compute scheduler from unit: 'nova-compute/1'
@@ -437,7 +437,7 @@ def test_hypervisor_upgrade_plan(model):
                     ├── Upgrade the unit: 'nova-compute/1'
                     ├── Resume the unit: 'nova-compute/1'
             Enable nova-compute scheduler from unit: 'nova-compute/1'
-            Wait for up to 1800s for model 'test_model' to reach the idle state
+            Wait for up to 2400s for model 'test_model' to reach the idle state
             Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/1
         Upgrade plan for 'az-2' to 'victoria'
             Disable nova-compute scheduler from unit: 'nova-compute/2'
@@ -454,7 +454,7 @@ def test_hypervisor_upgrade_plan(model):
                     ├── Upgrade the unit: 'nova-compute/2'
                     ├── Resume the unit: 'nova-compute/2'
             Enable nova-compute scheduler from unit: 'nova-compute/2'
-            Wait for up to 1800s for model 'test_model' to reach the idle state
+            Wait for up to 2400s for model 'test_model' to reach the idle state
             Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/2
     """
     )
@@ -547,7 +547,7 @@ def test_hypervisor_upgrade_plan_single_machine(model):
             Wait for up to 300s for app 'cinder' to reach the idle state
             Verify that the workload of 'cinder' has been upgraded on units: cinder/0
             Enable nova-compute scheduler from unit: 'nova-compute/0'
-            Wait for up to 1800s for model 'test_model' to reach the idle state
+            Wait for up to 2400s for model 'test_model' to reach the idle state
             Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/0
     """
     )
@@ -639,7 +639,7 @@ def test_hypervisor_upgrade_plan_some_units_upgraded(model):
             Wait for up to 300s for app 'cinder' to reach the idle state
             Verify that the workload of 'cinder' has been upgraded on units: cinder/2
             Enable nova-compute scheduler from unit: 'nova-compute/2'
-            Wait for up to 1800s for model 'test_model' to reach the idle state
+            Wait for up to 2400s for model 'test_model' to reach the idle state
             Verify that the workload of 'nova-compute' has been upgraded on units: nova-compute/2
     """
     )