sbcli-dev 6.3.4__tar.gz → 6.4.0__tar.gz

{sbcli_dev-6.3.4 → sbcli_dev-6.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sbcli-dev
-Version: 6.3.4
+Version: 6.4.0
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

sbcli_dev-6.4.0/env_var

@@ -0,0 +1,5 @@
+SIMPLY_BLOCK_COMMAND_NAME=sbcli-dev
+SIMPLY_BLOCK_VERSION=6.4.0
+
+SIMPLY_BLOCK_DOCKER_IMAGE=simplyblock/simplyblock:dev
+DOCKER_USER=hamdysimplyblock

{sbcli_dev-6.3.4 → sbcli_dev-6.4.0}/sbcli_dev.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sbcli-dev
-Version: 6.3.4
+Version: 6.4.0
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-6.3.4 → sbcli_dev-6.4.0}/simplyblock_core/constants.py

@@ -54,7 +54,7 @@ FDB_CHECK_INTERVAL_SEC = 60
 
 
 
-SIMPLY_BLOCK_DOCKER_IMAGE = "simplyblock/simplyblock:main"
+SIMPLY_BLOCK_DOCKER_IMAGE = "simplyblock/simplyblock:dev"
 SIMPLY_BLOCK_CLI_NAME = "sbcli-dev"
 TASK_EXEC_INTERVAL_SEC = 10
 TASK_EXEC_RETRY_COUNT = 8

{sbcli_dev-6.3.4 → sbcli_dev-6.4.0}/simplyblock_core/controllers/device_controller.py

@@ -632,6 +632,7 @@ def add_device(device_id):
         device_obj.status = NVMeDevice.STATUS_ONLINE
         snode.write_to_db(db_controller.kv_store)
         device_events.device_create(device_obj)
+        distr_controller.send_cluster_map_to_node(snode)
 
         logger.info("Make other nodes connect to the node devices")
         snodes = db_controller.get_storage_nodes_by_cluster_id(snode.cluster_id)
@@ -641,6 +642,8 @@ def add_device(device_id):
             node.remote_devices = storage_node_ops._connect_to_remote_devs(node)
             node.write_to_db()
 
+            distr_controller.send_cluster_map_to_node(node)
+
         tasks_controller.add_new_device_mig_task(device_id)
         return device_id
 
@@ -686,6 +689,7 @@ def add_device(device_id):
         logger.error("failed to create devices")
         return False
 
+    distr_controller.send_cluster_map_to_node(snode)
     logger.info("Make other nodes connect to the node devices")
     snodes = db_controller.get_storage_nodes_by_cluster_id(snode.cluster_id)
     for node in snodes:
@@ -693,6 +697,7 @@ def add_device(device_id):
             continue
         node.remote_devices = storage_node_ops._connect_to_remote_devs(node)
         node.write_to_db()
+        distr_controller.send_cluster_map_to_node(node)
 
     for dev in new_devices:
         tasks_controller.add_new_device_mig_task(dev.get_id())

{sbcli_dev-6.3.4 → sbcli_dev-6.4.0}/simplyblock_core/services/new_device_discovery.py

@@ -24,18 +24,19 @@ while True:
             logger.warning(f"Skipping node, id: {node.get_id()}, status: {node.status}")
             continue
 
-        known_pcie = [dev.pcie_address for dev in node.nvme_devices]
-        if node.jm_device and 'pcie_address' in node.jm_device.device_data_dict:
-            known_pcie.append(node.jm_device.device_data_dict['pcie_address'])
+        known_sn = [dev.serial_number for dev in node.nvme_devices]
+        if node.jm_device and 'serial_number' in node.jm_device.device_data_dict:
+            known_sn.append(node.jm_device.device_data_dict['serial_number'])
 
         snode_api = SNodeClient(node.api_endpoint)
         node_info, _ = snode_api.info()
 
-        node_pcie = node_info['spdk_pcie_list']
 
         # check for unused nvme devices
         if "lsblk" in node_info:
             for dev in node_info['nvme_devices']:
+                if dev['serial_number'] in known_sn:
+                    continue
                 for block_dev in node_info['lsblk']['blockdevices']:
                     if block_dev['name'] == dev['device_name']:
                         if 'children' not in block_dev:
@@ -43,18 +44,12 @@ while True:
                             # try mount to spdk driver
                             snode_api.bind_device_to_spdk(dev['address'])
                             time.sleep(3)
-
-        # check for dev again
-        node_info, _ = snode_api.info()
-        node_pcie = node_info['spdk_pcie_list']
-        for pc in node_pcie:
-            if pc not in known_pcie:
-                logger.info(f"New ssd found: {pc}")
-                devs = storage_node_ops.addNvmeDevices(node, [pc])
-                if devs:
-                    new_dev = devs[0]
-                    new_dev.status = NVMeDevice.STATUS_NEW
-                    node.nvme_devices.append(new_dev)
-                    node.write_to_db(db_controller.kv_store)
+                            devs = storage_node_ops.addNvmeDevices(node, [dev['address']])
+                            if devs:
+                                logger.info(f"New ssd found: {dev['address']}")
+                                new_dev = devs[0]
+                                new_dev.status = NVMeDevice.STATUS_NEW
+                                node.nvme_devices.append(new_dev)
+                                node.write_to_db(db_controller.kv_store)
 
     time.sleep(constants.DEV_DISCOVERY_INTERVAL_SEC)

{sbcli_dev-6.3.4 → sbcli_dev-6.4.0}/simplyblock_core/services/storage_node_monitor.py

@@ -296,11 +296,11 @@ while True:
             else:
                 set_node_offline(snode)
 
-                # if not ping_check and not node_api_check and not spdk_process:
-                #     # restart on new node
-                #     storage_node_ops.set_node_status(snode.get_id(), StorageNode.STATUS_SCHEDULABLE)
+                if not ping_check and not node_api_check and not spdk_process:
+                    # restart on new node
+                    storage_node_ops.set_node_status(snode.get_id(), StorageNode.STATUS_SCHEDULABLE)
 
-                if ping_check and node_api_check and (not spdk_process or not node_rpc_check):
+                elif ping_check and node_api_check and (not spdk_process or not node_rpc_check):
                     # add node to auto restart
                     if cluster.status in [Cluster.STATUS_ACTIVE, Cluster.STATUS_DEGRADED, Cluster.STATUS_SUSPENDED]:
                         tasks_controller.add_node_to_auto_restart(snode)

{sbcli_dev-6.3.4 → sbcli_dev-6.4.0}/simplyblock_core/storage_node_ops.py

@@ -446,7 +446,7 @@ def _create_storage_device_stack(rpc_client, nvme, snode, after_restart):
     alceml_name = device_controller.get_alceml_name(alceml_id)
     logger.info(f"adding {alceml_name}")
     pba_init_mode = 3
-    if after_restart:
+    if after_restart and nvme.status != NVMeDevice.STATUS_NEW:
         pba_init_mode = 1
     alceml_cpu_mask = ""
     alceml_worker_cpu_mask = ""
@@ -875,15 +875,15 @@ def _connect_to_remote_jm_devs(this_node, jm_ids=[]):
         if not org_dev:
             continue
 
-        name = f"remote_{jm_dev.jm_bdev}"
+        name = f"remote_{org_dev.jm_bdev}"
         bdev_name = f"{name}n1"
-        jm_dev.remote_bdev = bdev_name
+        org_dev.remote_bdev = bdev_name
 
         if org_dev.status == NVMeDevice.STATUS_ONLINE:
             if bdev_name in node_bdev_names:
                 logger.debug(f"bdev found {bdev_name}")
-                jm_dev.status = JMDevice.STATUS_ONLINE
-                new_devs.append(jm_dev)
+                org_dev.status = JMDevice.STATUS_ONLINE
+                new_devs.append(org_dev)
             else:
                 if rpc_client.bdev_nvme_controller_list(name):
                     logger.info(f"detaching {name} from {this_node.get_id()}")
@@ -892,21 +892,21 @@ def _connect_to_remote_jm_devs(this_node, jm_ids=[]):
 
                 logger.info(f"Connecting {name} to {this_node.get_id()}")
                 ret = rpc_client.bdev_nvme_attach_controller_tcp_JM(
-                    name, jm_dev.nvmf_nqn, jm_dev.nvmf_ip, jm_dev.nvmf_port)
+                    name, org_dev.nvmf_nqn, org_dev.nvmf_ip, org_dev.nvmf_port)
                 if ret:
-                    jm_dev.status = JMDevice.STATUS_ONLINE
+                    org_dev.status = JMDevice.STATUS_ONLINE
                 else:
                     logger.error(f"failed to connect to remote JM {name}")
-                    jm_dev.status = JMDevice.STATUS_UNAVAILABLE
-                new_devs.append(jm_dev)
+                    org_dev.status = JMDevice.STATUS_UNAVAILABLE
+                new_devs.append(org_dev)
 
         else:
             # if bdev_name in node_bdev_names:
             #     logger.debug(f"bdev found {bdev_name}")
             #     rpc_client.bdev_nvme_detach_controller(name)
 
-            jm_dev.status = JMDevice.STATUS_UNAVAILABLE
-            new_devs.append(jm_dev)
+            org_dev.status = JMDevice.STATUS_UNAVAILABLE
+            new_devs.append(org_dev)
 
     return new_devs
 
@@ -1829,6 +1829,9 @@ def restart_storage_node(
                         jm_device = nvme
                 jm_device.status = NVMeDevice.STATUS_JM
 
+                if snode.jm_device and snode.jm_device.get_id():
+                    jm_device.uuid = snode.jm_device.get_id()
+
                 ret = _prepare_cluster_devices_jm_on_dev(snode, nvme_devs)
             else:
                 ret = _prepare_cluster_devices_partitions(snode, nvme_devs)

sbcli_dev-6.3.4/env_var

@@ -1,5 +0,0 @@
-SIMPLY_BLOCK_COMMAND_NAME=sbcli-dev
-SIMPLY_BLOCK_VERSION=6.3.4
-
-SIMPLY_BLOCK_DOCKER_IMAGE=simplyblock/simplyblock:main
-DOCKER_USER=hamdysimplyblock