From hotfix multi fix to main (#738)

simplyblock_core/cluster_ops.py

@@ -458,7 +458,6 @@ def add_cluster(blk_size, page_size_in_blocks, cap_warn, cap_crit, prov_cap_warn
     cluster.strict_node_anti_affinity = strict_node_anti_affinity
 
     default_cluster = clusters[0]
-    cluster.mode = default_cluster.mode
     cluster.db_connection = default_cluster.db_connection
     cluster.grafana_secret = monitoring_secret if default_cluster.mode == "kubernetes" else default_cluster.grafana_secret
     cluster.grafana_endpoint = default_cluster.grafana_endpoint
@@ -1132,6 +1131,7 @@ def get_logs(cluster_id, limit=50, **kwargs) -> t.List[dict]:
         if record.event in ["device_status", "node_status"]:
             msg = msg+f" ({record.count})"
 
+        logger.debug(record)
         out.append({
             "Date": record.get_date_string(),
             "NodeId": record.node_id,
@@ -1154,10 +1154,6 @@ def update_cluster(cluster_id, mgmt_only=False, restart=False, spdk_image=None,
 
     logger.info("Updating mgmt cluster")
     if cluster.mode == "docker":
-        sbcli=constants.SIMPLY_BLOCK_CLI_NAME
-        subprocess.check_call(f"pip install {sbcli} --upgrade".split(' '))
-        logger.info(f"{sbcli} upgraded")
-
         cluster_docker = utils.get_docker_client(cluster_id)
         logger.info(f"Pulling image {constants.SIMPLY_BLOCK_DOCKER_IMAGE}")
         pull_docker_image_with_retry(cluster_docker, constants.SIMPLY_BLOCK_DOCKER_IMAGE)
@@ -1171,7 +1167,7 @@ def update_cluster(cluster_id, mgmt_only=False, restart=False, spdk_image=None,
         for service in cluster_docker.services.list():
             if image_parts in service.attrs['Spec']['Labels']['com.docker.stack.image'] or \
             "simplyblock" in service.attrs['Spec']['Labels']['com.docker.stack.image']:
-                if service.name == "app_CachingNodeMonitor":
+                if service.name in ["app_CachingNodeMonitor", "app_CachedLVolStatsCollector"]:
                     logger.info(f"Removing service {service.name}")
                     service.remove()
                 else:
@@ -1281,7 +1277,12 @@ def update_cluster(cluster_id, mgmt_only=False, restart=False, spdk_image=None,
                 logger.info(f"Restarting node: {node.get_id()} with SPDK image: {spdk_image}")
             else:
                 logger.info(f"Restarting node: {node.get_id()}")
-            storage_node_ops.restart_storage_node(node.get_id(), force=True, spdk_image=spdk_image)
+            try:
+                storage_node_ops.restart_storage_node(node.get_id(), force=True, spdk_image=spdk_image)
+            except Exception as e:
+                logger.debug(e)
+                logger.error(f"Failed to restart node: {node.get_id()}")
+                return
 
     logger.info("Done")
 

simplyblock_core/constants.py

@@ -139,7 +139,6 @@ def get_config_var(name, default=None):
 CLIENT_QPAIR_COUNT=3
 NVME_TIMEOUT_US=8000000
 NVMF_MAX_SUBSYSTEMS=50000
-HA_JM_COUNT=3
 KATO=10000
 ACK_TO=11
 BDEV_RETRY=0
@@ -225,4 +224,5 @@ def get_config_var(name, default=None):
 
 qos_class_meta_and_migration_weight_percent = 25
 
-MIG_PARALLEL_JOBS = 64
+MIG_PARALLEL_JOBS = 64
+MIG_JOB_SIZE = 64

simplyblock_core/controllers/device_controller.py

@@ -653,81 +653,6 @@ def add_device(device_id, add_migration_task=True):
         tasks_controller.add_new_device_mig_task(device_id)
     return device_id
 
-    #
-    # # create partitions
-    # partitions = snode.num_partitions_per_dev
-    # rpc_client = RPCClient(snode.mgmt_ip, snode.rpc_port, snode.rpc_username, snode.rpc_password)
-    # # look for partitions
-    # partitioned_devices = storage_node_ops._search_for_partitions(rpc_client, device_obj)
-    # logger.debug("partitioned_devices")
-    # logger.debug(partitioned_devices)
-    # if len(partitioned_devices) == partitions+1:
-    #     logger.info("Partitioned devices found")
-    # else:
-    #     logger.info(f"Creating partitions for {device_obj.nvme_bdev}")
-    #     storage_node_ops._create_device_partitions(rpc_client, device_obj, snode, partitions, snode.jm_percent)
-    #     partitioned_devices = storage_node_ops._search_for_partitions(rpc_client, device_obj)
-    #     if len(partitioned_devices) == partitions+1:
-    #         logger.info("Device partitions created")
-    #     else:
-    #         logger.error("Failed to create partitions")
-    #         return False
-    #
-    # jm_part = partitioned_devices.pop(0)
-    # new_devices = []
-    # dev_order = storage_node_ops.get_next_cluster_device_order(db_controller, snode.cluster_id)
-    # for dev in partitioned_devices:
-    #     new_device = storage_node_ops._create_storage_device_stack(rpc_client, dev, snode, after_restart=False)
-    #     if not new_device:
-    #         logger.error("failed to create dev stack")
-    #         continue
-    #
-    #     new_device.cluster_device_order = dev_order
-    #     dev_order += 1
-    #     device_events.device_create(new_device)
-    #     new_devices.append(new_device)
-    #
-    # if new_devices:
-    #     snode.nvme_devices.remove(device_obj)
-    #     snode.nvme_devices.extend(new_devices)
-    #     snode.write_to_db(db_controller.kv_store)
-    # else:
-    #     logger.error("failed to create devices")
-    #     return False
-    #
-    # for dev in new_devices:
-    #     distr_controller.send_cluster_map_add_device(dev, 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:
-    #     if node.get_id() == snode.get_id() or node.status != StorageNode.STATUS_ONLINE:
-    #         continue
-    #     node.remote_devices = storage_node_ops._connect_to_remote_devs(node)
-    #     node.write_to_db()
-    #     for dev in new_devices:
-    #         distr_controller.send_cluster_map_add_device(dev, node)
-    #
-    # for dev in new_devices:
-    #     tasks_controller.add_new_device_mig_task(dev.get_id())
-    #
-    # # add to jm raid
-    # if snode.jm_device and snode.jm_device.raid_bdev and jm_part:
-    #     # looking for jm partition
-    #     jm_dev_part = jm_part.nvme_bdev
-    #     ret = rpc_client.get_bdevs(jm_dev_part)
-    #     if ret:
-    #         logger.info(f"JM part found: {jm_dev_part}")
-    #         if snode.jm_device.status in [JMDevice.STATUS_UNAVAILABLE, JMDevice.STATUS_REMOVED]:
-    #             restart_jm_device(snode.jm_device.get_id(), force=True, format_alceml=True)
-    #
-    #         if snode.jm_device.status == JMDevice.STATUS_ONLINE and \
-    #                 jm_dev_part not in snode.jm_device.jm_nvme_bdev_list:
-    #             remove_jm_device(snode.jm_device.get_id(), force=True)
-    #             restart_jm_device(snode.jm_device.get_id(), force=True)
-    #
-    # return "Done"
-
 
 def device_set_failed_and_migrated(device_id):
     db_controller = DBController()