test(robot): add drain cases from manual test cases

longhorn/longhorn-9292

Signed-off-by: Chris <[email protected]>

e2e/keywords/common.resource

@@ -19,6 +19,7 @@ Library             ../libs/keywords/setting_keywords.py
 Library             ../libs/keywords/backupstore_keywords.py
 Library             ../libs/keywords/backup_keywords.py
 Library             ../libs/keywords/sharemanager_keywords.py
+Library             ../libs/keywords/k8s_keywords.py
 
 *** Keywords ***
 Set test environment
@@ -35,6 +36,7 @@ Set test environment
     END
 
 Cleanup test resources
+    uncordon_all_nodes
     cleanup_control_plane_network_latency
     reset_node_schedule
     cleanup_node_exec

e2e/keywords/host.resource

@@ -37,3 +37,8 @@ Restart cluster
 Power on off node
     Run keyword And Continue On Failure
     ...    power_on_node_by_name    ${powered_off_node}
+
+Power off node ${node_id}
+    ${powered_off_node} =     get_node_by_index    ${node_id}
+    power_off_node_by_name    ${powered_off_node}
+    Set Test Variable    ${powered_off_node}

e2e/keywords/k8s.resource

@@ -5,9 +5,10 @@ Library             ../libs/keywords/k8s_keywords.py
 Library             ../libs/keywords/workload_keywords.py
 Library             ../libs/keywords/volume_keywords.py
 Library             ../libs/keywords/host_keywords.py
+Library             ../libs/keywords/node_keywords.py
 
 *** Variables ***
-
+${DRAIN_TIMEOUT}    90
 
 *** Keywords ***
 Stop volume node kubelet of ${workload_kind} ${workload_id} for ${duration} seconds
@@ -51,5 +52,63 @@ Force drain volume of ${workload_kind} ${workload_id} replica node
     Set Test Variable    ${drained_node}
     Set Test Variable    ${last_volume_node}
 
+Force drain node ${node_id}
+    ${node_name} =    get_node_by_index    ${node_id}
+    force_drain_node    ${node_name}
+
+Drain volume of ${workload_kind} ${workload_id} volume node
+    ${workload_name} =   generate_name_with_suffix    ${workload_kind}    ${workload_id}
+    ${volume_name} =    get_workload_volume_name    ${workload_name}
+    ${drained_node} =    get_volume_node    ${volume_name}
+    ${last_volume_node} =    get_volume_node    ${volume_name}
+    drain_node    ${drained_node}
+    wait_for_all_pods_evicted    ${drained_node}
+    Set Test Variable    ${drained_node}
+    Set Test Variable    ${last_volume_node}
+
 Uncordon the drained node
     uncordon_node    ${drained_node}
+
+Cordon ${workload_kind} ${workload_id} volume node
+    ${workload_name} =   generate_name_with_suffix    ${workload_kind}    ${workload_id}
+    ${volume_name} =    get_workload_volume_name    ${workload_name}
+    ${volume_node} =    get_volume_node    ${volume_name}
+    cordon_node    ${volume_node}
+    check_node_is_not_schedulable    ${volume_node}
+
+Force drain all nodes
+    FOR    ${node_id}    IN RANGE    0    3
+        ${node_name} =    get_node_by_index    ${node_id}
+        force_drain_node    ${node_name}
+        wait_for_all_pods_evicted    ${node_name}
+    END
+
+Check node ${node_id} cordoned
+    ${node_name} =    get_node_by_index    ${node_id}
+    check_node_cordoned    ${node_name}
+
+Force drain node ${node_id} and expect failure
+    ${drained_node} =    get_node_by_index    ${node_id}
+    ${instance_manager_name} =     get_instance_manager_on_node    ${drained_node}
+    Run Keyword And Expect Error    *    force_drain_node    ${drained_node}
+    Set Test Variable    ${instance_manager_name}
+    Set Test Variable    ${drained_node}
+
+Force drain node ${node_id} and expect success
+    ${drained_node} =    get_node_by_index    ${node_id}
+    ${instance_manager_name} =     get_instance_manager_on_node    ${drained_node}
+    force_drain_node    ${drained_node}
+    Set Test Variable    ${instance_manager_name}
+    Set Test Variable    ${drained_node}
+
+The drain process not completed
+    check_drain_process_not_completed    ${drain_process}
+
+The drain process completed
+    wait_for_all_pods_evicted    ${drained_node}
+    check_drain_process_completed    ${drain_process}
+
+Check PDB not exist
+    [Arguments]    ${instance_manger}
+    check_instance_manager_pdb_not_exist    ${instance_manger}
+

e2e/keywords/longhorn.resource

@@ -3,6 +3,7 @@ Documentation       Longhorn Keywords
 
 Library             ../libs/keywords/instancemanager_keywords.py
 Library             ../libs/keywords/workload_keywords.py
+Library             ../libs/keywords/k8s_keywords.py
 
 *** Variables ***
 @{longhorn_workloads}
@@ -43,3 +44,12 @@ Check Longhorn workload pods ${condition} annotated with ${key}
     Run Keyword IF    '${condition}' == 'not'    Should Not Be True    ${is_annotated}
     ...    ELSE IF    '${condition}' == 'is'    Should Be True    ${is_annotated}
     ...    ELSE    Fail    Invalid condition ${condition}
+
+Check instance-manager pod is not running on drained node
+    ${pod} =     get_instance_manager_on_node    ${drained_node}
+    Should Be Equal    ${pod}    ${None}
+
+Check instance-manager pod is running on node ${node_id}
+    ${node_name} =    get_node_by_index    ${node_id}
+    ${pod} =     get_instance_manager_on_node    ${node_name}
+    Should Not Be Equal    ${pod}    ${None}

e2e/keywords/volume.resource

@@ -206,6 +206,12 @@ Wait volume ${volume_id} replica on node ${node_id} stopped
     ${node_name} =    get_node_by_index    ${node_id}
     wait_for_replica_stopped    ${volume_name}    ${node_name}
 
+Check volume ${volume_id} replica on node ${node_id} exist
+    ${volume_name} =    generate_name_with_suffix    volume    ${volume_id}
+    ${node_name} =    get_node_by_index    ${node_id}
+    ${replica_name}    get_replica_name_on_node    ${volume_name}    ${node_name}
+    Should Not Be Equal    ${replica_name}    ${None}
+
 Check volume ${volume_id} data is intact
     ${volume_name} =    generate_name_with_suffix    volume    ${volume_id}
     check_data_checksum    ${volume_name}

e2e/libs/k8s/k8s.py

@@ -1,6 +1,7 @@
 import time
 import subprocess
 import asyncio
+import os
 from kubernetes import client
 from workload.pod import create_pod
 from workload.pod import delete_pod
@@ -9,6 +10,8 @@
 from utility.utility import subprocess_exec_cmd
 from utility.utility import logging
 from utility.utility import get_retry_count_and_interval
+from utility.utility import subprocess_exec_cmd_with_timeout
+from robot.libraries.BuiltIn import BuiltIn
 
 async def restart_kubelet(node_name, downtime_in_sec=10):
     manifest = new_pod_manifest(
@@ -32,9 +35,9 @@ def drain_node(node_name):
     exec_cmd = ["kubectl", "drain", node_name, "--ignore-daemonsets", "--delete-emptydir-data"]
     res = subprocess_exec_cmd(exec_cmd)
 
-def force_drain_node(node_name):
+def force_drain_node(node_name, timeout):
     exec_cmd = ["kubectl", "drain", node_name, "--force", "--ignore-daemonsets", "--delete-emptydir-data"]
-    res = subprocess_exec_cmd(exec_cmd)
+    res = subprocess_exec_cmd_with_timeout(exec_cmd, timeout)
 
 def cordon_node(node_name):
     exec_cmd = ["kubectl", "cordon", node_name]
@@ -71,3 +74,23 @@ def wait_all_pods_evicted(node_name):
         time.sleep(retry_interval)
 
     assert evicted, 'failed to evict pods'
+
+def check_node_cordoned(node_name):
+    api = client.CoreV1Api()
+    node = api.read_node(node_name)
+    assert node.spec.unschedulable is True, f"node {node_name} is not cordoned."
+
+def get_instance_manager_on_node(node_name):
+    data_engine = BuiltIn().get_variable_value("${DATA_ENGINE}")
+    pods = get_all_pods_on_node(node_name)
+    for pod in pods:
+        labels = pod.metadata.labels
+        if labels.get("longhorn.io/data-engine") == data_engine and \
+           labels.get("longhorn.io/component") == "instance-manager":
+            return pod.metadata.name
+    return None
+
+def check_instance_manager_pdb_not_exist(instance_manager):
+    exec_cmd = ["kubectl", "get", "pdb", "-n", "longhorn-system"]
+    res = subprocess_exec_cmd(exec_cmd)
+    assert instance_manager not in res.decode('utf-8')

e2e/libs/keywords/host_keywords.py

@@ -52,3 +52,6 @@ def power_off_volume_node(self, volume_name):
 
     def power_on_node_by_name(self, node_name):
         self.host.power_on_node(node_name)
+
+    def power_off_node_by_name(self, node_name):
+        self.host.power_off_node(node_name)

e2e/libs/keywords/k8s_keywords.py

@@ -5,8 +5,12 @@
 from k8s.k8s import drain_node, force_drain_node
 from k8s.k8s import cordon_node, uncordon_node
 from k8s.k8s import wait_all_pods_evicted
+from k8s.k8s import get_all_pods_on_node
+from k8s.k8s import check_node_cordoned
+from k8s.k8s import get_instance_manager_on_node
+from k8s.k8s import check_instance_manager_pdb_not_exist
 from utility.utility import logging
-
+from node import Node
 
 class k8s_keywords:
 
@@ -45,10 +49,32 @@ def drain_node(self, node_name):
         drain_node(node_name)
 
     def force_drain_node(self, node_name):
-        force_drain_node(node_name)
+        timeout = int(BuiltIn().get_variable_value("${DRAIN_TIMEOUT}", default="90"))
+        force_drain_node(node_name, timeout)
 
     def uncordon_node(self, node_name):
         uncordon_node(node_name)
 
+    def cordon_node(self, node_name):
+        cordon_node(node_name)
+
     def wait_for_all_pods_evicted(self, node_name):
         wait_all_pods_evicted(node_name)
+
+    def uncordon_all_nodes(self):
+        nodes = Node.list_node_names_by_role("worker")
+
+        for node_name in nodes:
+            uncordon_node(node_name)
+
+    def get_all_pods_on_node(self, node_name):
+        return get_all_pods_on_node(node_name)
+
+    def check_node_cordoned(self, node_name):
+        check_node_cordoned(node_name)
+
+    def get_instance_manager_on_node(self, node_name):
+        return get_instance_manager_on_node(node_name)
+
+    def check_instance_manager_pdb_not_exist(self, instance_manager):
+        return check_instance_manager_pdb_not_exist(instance_manager)

e2e/libs/keywords/node_keywords.py

@@ -40,5 +40,9 @@ def enable_node_scheduling(self, node_name):
 
     def reset_node_schedule(self):
         nodes = self.node.list_node_names_by_role("worker")
+
         for node_name in nodes:
             self.enable_node_scheduling(node_name)
+
+    def check_node_is_not_schedulable(self, node_name):
+        self.node.check_node_schedulable(node_name, schedulable="False")

e2e/libs/node/node.py

@@ -10,6 +10,8 @@
 from utility.utility import get_retry_count_and_interval
 from utility.utility import logging
 
+from k8s.k8s import uncordon_node
+
 class Node:
 
     DEFAULT_DISK_PATH = "/var/lib/longhorn/"
@@ -159,3 +161,11 @@ def set_default_disk_scheduling(self, node_name, allowScheduling):
             if disk.path == self.DEFAULT_DISK_PATH:
                 disk.allowScheduling = allowScheduling
         self.update_disks(node_name, node.disks)
+
+    def check_node_schedulable(self, node_name, schedulable):
+        node = get_longhorn_client().by_id_node(node_name)
+        for _ in range(self.retry_count):
+            if node["conditions"]["Schedulable"]["status"] == schedulable:
+                break
+            time.sleep(self.retry_interval)
+        assert node["conditions"]["Schedulable"]["status"] == schedulable

e2e/libs/utility/utility.py

@@ -86,6 +86,10 @@ def subprocess_exec_cmd(cmd):
     logging(f"Executed command {cmd} with result {res}")
     return res
 
+def subprocess_exec_cmd_with_timeout(cmd, timeout):
+    res = subprocess.check_output(cmd, timeout=timeout)
+    logging(f"Executed command {cmd} with timeout {timeout}s, result {res}")
+    return res
 
 def wait_for_cluster_ready():
     core_api = client.CoreV1Api()