sbcli-dev 10.4.1__tar.gz → 10.4.3__tar.gz

{sbcli_dev-10.4.1 → sbcli_dev-10.4.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: sbcli-dev
-Version: 10.4.1
+Version: 10.4.3
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-10.4.1 → sbcli_dev-10.4.3}/sbcli_dev.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: sbcli-dev
-Version: 10.4.1
+Version: 10.4.3
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-10.4.1 → sbcli_dev-10.4.3}/simplyblock_core/constants.py

@@ -73,7 +73,7 @@ SIMPLY_BLOCK_CLI_NAME = get_from_env_var_file(
         "SIMPLY_BLOCK_COMMAND_NAME", "sbcli")
 TASK_EXEC_INTERVAL_SEC = 10
 TASK_EXEC_RETRY_COUNT = 8
-SIMPLY_BLOCK_SPDK_CORE_IMAGE = "simplyblock/spdk-core:v24.05.x-latest"
+SIMPLY_BLOCK_SPDK_CORE_IMAGE = "simplyblock/spdk-core:v24.05-tag-latest"
 
 
 SIMPLY_BLOCK_SPDK_ULTRA_IMAGE = "simplyblock/spdk:lvolstore-ha-snapshot-latest"

{sbcli_dev-10.4.1 → sbcli_dev-10.4.3}/simplyblock_core/controllers/caching_node_controller.py

@@ -176,7 +176,12 @@ def add_node(cluster_id, node_ip, iface_name, data_nics_list, spdk_cpu_mask, spd
         logger.error("Pod is not running, exiting")
         return False
 
-    time.sleep(20)
+    time.sleep(3)
+
+    snode.namespace = namespace
+    snode.spdk_cpu_mask = spdk_cpu_mask
+    snode.spdk_mem = spdk_mem
+    snode.spdk_image = spdk_image
 
     # creating RPCClient instance
     rpc_client = RPCClient(
@@ -216,50 +221,50 @@ def add_node(cluster_id, node_ip, iface_name, data_nics_list, spdk_cpu_mask, spd
 
     cache_size = 0
     cache_bdev = None
-    if supported_ssd_size < ssd_size:
-        logger.info(f"SSD size is bigger than the supported size, will use split bdev: {split_factor}")
-        ret = rpc_client.bdev_split(ssd_dev.nvme_bdev, split_factor)
-        cache_bdev = ret[0]
-        cache_size = int(ssd_dev.size/split_factor)
-        snode.cache_split_factor = split_factor
-    else:
-        snode.cache_split_factor = 0
-        cache_bdev = ssd_dev.nvme_bdev
-        cache_size = ssd_dev.size
-
     # if supported_ssd_size < ssd_size:
-    #     logger.info(f"SSD size is bigger than the supported size, creating partition")
-    #
-    #     nbd_device = rpc_client.nbd_start_disk(ssd_dev.nvme_bdev)
-    #     time.sleep(3)
-    #     if not nbd_device:
-    #         logger.error(f"Failed to start nbd dev")
-    #         return False
-    #
-    #     jm_percent = int((supported_ssd_size/ssd_size) * 100)
-    #     result, error = cnode_api.make_gpt_partitions(nbd_device, jm_percent)
-    #     if error:
-    #         logger.error(f"Failed to make partitions")
-    #         logger.error(error)
-    #         return False
-    #     time.sleep(3)
-    #     rpc_client.nbd_stop_disk(nbd_device)
-    #     time.sleep(1)
-    #     rpc_client.bdev_nvme_detach_controller(ssd_dev.nvme_controller)
-    #     time.sleep(1)
-    #     rpc_client.bdev_nvme_controller_attach(ssd_dev.nvme_controller, ssd_dev.pcie_address)
-    #     time.sleep(1)
-    #     rpc_client.bdev_examine(ssd_dev.nvme_bdev)
-    #     time.sleep(1)
-    #
-    #     cache_bdev = f"{ssd_dev.nvme_bdev}p1"
-    #     cache_size = int(supported_ssd_size)
-    #
+    #     logger.info(f"SSD size is bigger than the supported size, will use split bdev: {split_factor}")
+    #     ret = rpc_client.bdev_split(ssd_dev.nvme_bdev, split_factor)
+    #     cache_bdev = ret[0]
+    #     cache_size = int(ssd_dev.size/split_factor)
+    #     snode.cache_split_factor = split_factor
     # else:
-    #
+    #     snode.cache_split_factor = 0
     #     cache_bdev = ssd_dev.nvme_bdev
     #     cache_size = ssd_dev.size
 
+    if supported_ssd_size < ssd_size:
+        logger.info(f"SSD size is bigger than the supported size, creating partition")
+
+        nbd_device = rpc_client.nbd_start_disk(ssd_dev.nvme_bdev)
+        time.sleep(3)
+        if not nbd_device:
+            logger.error(f"Failed to start nbd dev")
+            return False
+
+        jm_percent = int((supported_ssd_size/ssd_size) * 100)
+        result, error = cnode_api.make_gpt_partitions(nbd_device, jm_percent)
+        if error:
+            logger.error(f"Failed to make partitions")
+            logger.error(error)
+            return False
+        time.sleep(3)
+        rpc_client.nbd_stop_disk(nbd_device)
+        time.sleep(1)
+        rpc_client.bdev_nvme_detach_controller(ssd_dev.nvme_controller)
+        time.sleep(1)
+        rpc_client.bdev_nvme_controller_attach(ssd_dev.nvme_controller, ssd_dev.pcie_address)
+        time.sleep(1)
+        rpc_client.bdev_examine(ssd_dev.nvme_bdev)
+        time.sleep(1)
+
+        cache_bdev = f"{ssd_dev.nvme_bdev}p1"
+        cache_size = int(supported_ssd_size)
+
+    else:
+
+        cache_bdev = ssd_dev.nvme_bdev
+        cache_size = ssd_dev.size
+
     logger.info(f"Cache size: {utils.humanbytes(cache_size)}")
 
     snode.cache_bdev = cache_bdev
@@ -297,7 +302,32 @@ def recreate(node_id):
         return False
 
     logger.info(f"Recreating caching node: {node_id}, status: {snode.status}")
-    snode_api = CNodeClient(f"{snode.mgmt_ip}:5000")
+    cnode_api = CNodeClient(f"{snode.mgmt_ip}:5000")
+
+    if not cnode_api.spdk_process_is_up():
+        results, err = cnode_api.spdk_process_start(
+            snode.spdk_cpu_mask, snode.spdk_mem, snode.spdk_image, snode.mgmt_ip,
+            snode.rpc_port, snode.rpc_username, snode.rpc_password, snode.namespace)
+        if not results:
+            logger.error(f"Failed to start spdk: {err}")
+            return False
+
+        retries = 20
+        while retries > 0:
+            resp, _ = cnode_api.spdk_process_is_up()
+            if resp:
+                logger.info(f"Pod is up")
+                break
+            else:
+                logger.info("Pod is not running, waiting...")
+                time.sleep(3)
+                retries -= 1
+
+        if retries == 0:
+            logger.error("Pod is not running, exiting")
+            return False
+
+        time.sleep(5)
 
     # creating RPCClient instance
     rpc_client = RPCClient(
@@ -306,7 +336,7 @@ def recreate(node_id):
         timeout=60*5, retry=5)
 
     # get new node info after starting spdk
-    node_info, _ = snode_api.info()
+    node_info, _ = cnode_api.info()
     # adding devices
     nvme_devs = addNvmeDevices(rpc_client, node_info['spdk_pcie_list'], snode)
     if not nvme_devs:
@@ -460,7 +490,7 @@ def connect(caching_node_id, lvol_id):
             ret, _ = cnode_client.connect_nvme(ip, "4420", subsystem_nqn)
             break
 
-    time.sleep(5)
+    time.sleep(3)
     cnode_info, _ = cnode_client.info()
     nvme_devs = cnode_info['nvme_devices']
     dev_path = None
@@ -625,8 +655,8 @@ def deploy(ifname):
             f"DOCKER_IP={dev_ip}"
         ]
     )
-    logger.info("Pulling SPDK images")
-    node_docker.images.pull(constants.SIMPLY_BLOCK_DOCKER_IMAGE)
+    logger.info(f"Pulling SPDK image {constants.SIMPLY_BLOCK_SPDK_CORE_IMAGE}")
+    node_docker.images.pull(constants.SIMPLY_BLOCK_SPDK_CORE_IMAGE)
     return f"{dev_ip}:5000"
 
 

{sbcli_dev-10.4.1 → sbcli_dev-10.4.3}/simplyblock_core/env_var

@@ -1,5 +1,5 @@
 SIMPLY_BLOCK_COMMAND_NAME=sbcli-dev
-SIMPLY_BLOCK_VERSION=10.4.1
+SIMPLY_BLOCK_VERSION=10.4.3
 
 SIMPLY_BLOCK_DOCKER_IMAGE=simplyblock/simplyblock:main
 DOCKER_USER=hamdysimplyblock

{sbcli_dev-10.4.1 → sbcli_dev-10.4.3}/simplyblock_core/models/caching_node.py

@@ -50,3 +50,7 @@ class CachingNode(BaseNodeObject):
     services: List[str] = []
     subsystem: str = ""
     system_uuid: str = ""
+    namespace: str = ""
+    spdk_cpu_mask: str = ""
+    spdk_mem: int = ""
+    spdk_image: str = ""

{sbcli_dev-10.4.1 → sbcli_dev-10.4.3}/simplyblock_web/blueprints/caching_node_ops.py

@@ -295,17 +295,20 @@ def get_huge_memory():
         return 0
 
 
+CPU_INFO = cpuinfo.get_cpu_info()
+HOSTNAME, _, _ = node_utils.run_command("hostname -s")
+SYSTEM_ID, _, _ = node_utils.run_command("dmidecode -s system-uuid")
+
+
 @bp.route('/info', methods=['GET'])
 def get_info():
-    hostname, _, _ = run_command("hostname -s")
-    system_id, _, _ = run_command("dmidecode -s system-uuid")
 
     out = {
-        "hostname": hostname,
-        "system_id": system_id,
+        "hostname": HOSTNAME,
+        "system_id": SYSTEM_ID,
 
-        "cpu_count": cpuinfo.get_cpu_info()['count'],
-        "cpu_hz": cpuinfo.get_cpu_info()['hz_advertised'][0] if 'hz_advertised' in cpuinfo.get_cpu_info() else 1,
+        "cpu_count": CPU_INFO['count'],
+        "cpu_hz": CPU_INFO['hz_advertised'][0] if 'hz_advertised' in CPU_INFO else 1,
 
         "memory": get_memory(),
         "hugepages": get_huge_memory(),