sbcli-dev 4.0.12__zip → 4.0.14__zip

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sbcli-dev
-Version: 4.0.12
+Version: 4.0.14
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/env_var

@@ -1,5 +1,5 @@
 SIMPLY_BLOCK_COMMAND_NAME=sbcli-dev
-SIMPLY_BLOCK_VERSION=4.0.12
+SIMPLY_BLOCK_VERSION=4.0.14
 
 
 SIMPLY_BLOCK_DOCKER_IMAGE=simplyblock/simplyblock:main-dev

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/sbcli_dev.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sbcli-dev
-Version: 4.0.12
+Version: 4.0.14
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/simplyblock_cli/cli.py

@@ -1264,7 +1264,6 @@ class CLIWrapper:
         contact_point = args.contact_point
         grafana_endpoint = args.grafana_endpoint
         enable_node_affinity = args.enable_node_affinity
-        enable_node_affinity = True
 
         return cluster_ops.create_cluster(
             blk_size, page_size_in_blocks,

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/simplyblock_core/services/main_distr_event_collector.py

@@ -33,17 +33,8 @@ def process_device_event(event):
                     if dev.status not in [NVMeDevice.STATUS_ONLINE, NVMeDevice.STATUS_READONLY]:
                         logger.info(f"The storage device is not online, skipping. status: {dev.status}")
                         event.status = 'skipped'
-                        node_status_event = {
-                            "timestamp": datetime.datetime.now().isoformat("T", "seconds") + 'Z',
-                            "event_type": "device_status",
-                            "storage_ID": storage_id,
-                            "status": dev.status}
-                        events = {"events": [node_status_event]}
-                        snode = db_controller.get_storage_node_by_id(node_id)
-                        rpc_client = RPCClient(snode.mgmt_ip, snode.rpc_port, snode.rpc_username, snode.rpc_password)
-                        ret = rpc_client.distr_status_events_update(events)
-                        if not ret:
-                            logger.warning("Failed to send event update")
+                        n = db_controller.get_storage_node_by_id(node_id)
+                        distr_controller.send_dev_status_event(dev, NVMeDevice.STATUS_UNAVAILABLE, n)
                         return
 
                     device = dev
@@ -57,30 +48,9 @@ def process_device_event(event):
         device_id = device.get_id()
         node = db_controller.get_storage_node_by_id(node_id)
         if device.node_id != node_id:
-            logger.info(f"Removing remote storage id: {storage_id} from node: {node_id}")
-            new_remote_devices = []
-            rpc_client = RPCClient(node.mgmt_ip, node.rpc_port,
-                                   node.rpc_username, node.rpc_password)
-            for rem_dev in node.remote_devices:
-                if rem_dev.get_id() == device.get_id():
-                    ctrl_name = rem_dev.remote_bdev[:-2]
-                    rpc_client.bdev_nvme_detach_controller(ctrl_name)
-                else:
-                    new_remote_devices.append(rem_dev)
-            node.remote_devices = new_remote_devices
-            node.write_to_db(db_controller.kv_store)
-            node_status_event = {
-                "timestamp": datetime.datetime.now().isoformat("T", "seconds") + 'Z',
-                "event_type": "device_status",
-                "storage_ID": storage_id,
-                "status": NVMeDevice.STATUS_UNAVAILABLE}
-            events = {"events": [node_status_event]}
-            ret = rpc_client.distr_status_events_update(events)
-            if not ret:
-                logger.warning("Failed to send event update")
-
-            dev_node = db_controller.get_storage_node_by_id(device.node_id)
-            if dev_node.status == StorageNode.STATUS_ONLINE:
+            logger.info(f"Setting storage id: {storage_id} unavailable")
+            distr_controller.send_dev_status_event(device, NVMeDevice.STATUS_UNAVAILABLE, node)
+            if device.status == NVMeDevice.STATUS_ONLINE:
                 device_controller.device_set_io_error(device_id, True)
                 device_controller.device_set_unavailable(device_id)
 
@@ -156,7 +126,7 @@ def start_event_collector_on_node(node_id):
         snode.rpc_port,
         snode.rpc_username,
         snode.rpc_password,
-        timeout=10, retry=2)
+        timeout=5, retry=2)
 
     while True:
         page = 1

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/simplyblock_core/services/storage_node_monitor.py

@@ -100,6 +100,16 @@ def update_cluster_status(cluster_id):
 
 def set_node_online(node):
     if node.status != StorageNode.STATUS_ONLINE:
+        # set devices online
+        for dev in node.nvme_devices:
+            if dev.status == NVMeDevice.STATUS_UNAVAILABLE:
+                device_controller.device_set_online(dev.get_id())
+
+        # set jm dev online
+        if snode.jm_device.status == JMDevice.STATUS_UNAVAILABLE:
+            device_controller.set_jm_device_state(snode.jm_device.get_id(), JMDevice.STATUS_ONLINE)
+
+        # set node online
         storage_node_ops.set_node_status(node.get_id(), StorageNode.STATUS_ONLINE)
 
 
@@ -110,7 +120,8 @@ def set_node_offline(node):
 
         # set devices unavailable
         for dev in node.nvme_devices:
-            device_controller.device_set_unavailable(dev.get_id())
+            if dev.status == NVMeDevice.STATUS_ONLINE:
+                device_controller.device_set_unavailable(dev.get_id())
 
         # set jm dev offline
         if snode.jm_device.status != JMDevice.STATUS_UNAVAILABLE:

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/simplyblock_core/services/tasks_runner_restart.py

@@ -178,7 +178,7 @@ def task_runner_node(task):
     ret = storage_node_ops.shutdown_storage_node(node.get_id(), force=True)
     if ret:
         logger.info(f"Node shutdown succeeded")
-    time.sleep(60)
+    time.sleep(3)
 
     # resetting node
     logger.info(f"Restart node {node.get_id()}")
@@ -219,4 +219,7 @@ while True:
                             tasks_events.task_updated(task)
                         else:
                             time.sleep(delay_seconds)
-                            delay_seconds *= 2
+                            if task.retry <= 3 and task.function_name == JobSchedule.FN_DEV_RESTART:
+                                delay_seconds *= 1
+                            else:
+                                delay_seconds *= 2

{sbcli_dev-4.0.12 → sbcli_dev-4.0.14}/simplyblock_web/blueprints/web_api_storage_node.py

@@ -197,6 +197,7 @@ def storage_node_add():
     max_prov = req_data['max_prov']
     number_of_distribs = int(req_data.get('number_of_distribs', 4))
     disable_ha_jm = bool(req_data.get('disable_ha_jm', False))
+    enable_test_device = bool(req_data.get('enable_test_device', False))
 
     spdk_image = None
     if 'spdk_image' in req_data:
@@ -259,7 +260,7 @@ def storage_node_add():
         "num_partitions_per_dev": partitions,
         "jm_percent": jm_percent,
         "number_of_devices": number_of_devices,
-        "enable_test_device": False,
+        "enable_test_device": enable_test_device,
         "number_of_distribs": number_of_distribs,
         "namespace": namespace,
         "enable_ha_jm": not disable_ha_jm})