sbcli-dev 6.3.0__tar.gz → 6.3.2__tar.gz

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sbcli-dev
-Version: 6.3.0
+Version: 6.3.2
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/env_var

@@ -1,5 +1,5 @@
 SIMPLY_BLOCK_COMMAND_NAME=sbcli-dev
-SIMPLY_BLOCK_VERSION=6.3.0
+SIMPLY_BLOCK_VERSION=6.3.2
 
 SIMPLY_BLOCK_DOCKER_IMAGE=simplyblock/simplyblock:main
 DOCKER_USER=hamdysimplyblock

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/sbcli_dev.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sbcli-dev
-Version: 6.3.0
+Version: 6.3.2
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/simplyblock_core/constants.py

@@ -108,5 +108,6 @@ SPDK_PROXY_MULTI_THREADING_ENABLED=True
 SPDK_PROXY_TIMEOUT=60*5
 LVOL_NVME_CONNECT_RECONNECT_DELAY=3
 LVOL_NVME_CONNECT_CTRL_LOSS_TMO=-1
-QPAIR_COUNT=20
+LVOL_NVME_CONNECT_NR_IO_QUEUES=6
+QPAIR_COUNT=32
 NVME_TIMEOUT_US=20000000

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/simplyblock_core/controllers/health_controller.py

@@ -276,7 +276,8 @@ def check_node(node_id, with_devices=True):
 
             if snode.secondary_node_id:
                 second_node_1 = db_controller.get_storage_node_by_id(snode.secondary_node_id)
-                lvstore_check &= _check_node_lvstore(lvstore_stack, second_node_1)
+                if second_node_1.status == StorageNode.STATUS_ONLINE:
+                    lvstore_check &= _check_node_lvstore(lvstore_stack, second_node_1)
 
         if snode.is_secondary_node:
             for node in db_controller.get_storage_nodes():

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/simplyblock_core/controllers/lvol_controller.py

@@ -1099,9 +1099,11 @@ def connect_lvol(uuid):
                 "nqn": lvol.nqn,
                 "reconnect-delay": constants.LVOL_NVME_CONNECT_RECONNECT_DELAY,
                 "ctrl-loss-tmo": constants.LVOL_NVME_CONNECT_CTRL_LOSS_TMO,
+                "nr-io-queues": constants.LVOL_NVME_CONNECT_NR_IO_QUEUES,
                 "connect": f"sudo nvme connect --reconnect-delay={constants.LVOL_NVME_CONNECT_RECONNECT_DELAY} "
-                           f"--ctrl-loss-tmo={constants.LVOL_NVME_CONNECT_CTRL_LOSS_TMO} --transport={transport} "
-                           f"--traddr={ip} --trsvcid={port} --nqn={lvol.nqn}",
+                           f"--ctrl-loss-tmo={constants.LVOL_NVME_CONNECT_CTRL_LOSS_TMO} "
+                           f"--nr-io-queues={constants.LVOL_NVME_CONNECT_NR_IO_QUEUES} "
+                           f"--transport={transport} --traddr={ip} --trsvcid={port} --nqn={lvol.nqn}",
             })
     return out
 

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/simplyblock_core/services/lvol_stat_collector.py

@@ -1,10 +1,10 @@
 # coding=utf-8
 import time
-from email.policy import default
 
 from simplyblock_core import constants, kv_store, utils
 from simplyblock_core.controllers import lvol_events
 from simplyblock_core.models.stats import LVolStatObject, PoolStatObject
+from simplyblock_core.models.storage_node import StorageNode
 from simplyblock_core.rpc_client import RPCClient
 
 
@@ -210,12 +210,13 @@ while True:
                 hosts=[lvol.node_id]
             for host in hosts:
                 snode = db_controller.get_storage_node_by_id(host)
+                if snode.status != StorageNode.STATUS_ONLINE:
+                    continue
                 rpc_client = RPCClient(
                     snode.mgmt_ip, snode.rpc_port,
                     snode.rpc_username, snode.rpc_password,
                     timeout=1, retry=2)
-
-                logger.info("Getting lVol stats: %s", lvol.uuid)
+                logger.info("Getting lVol stats: %s from node: %s", lvol.uuid, host)
                 stats_dict = rpc_client.get_lvol_stats(lvol.top_bdev)
                 if stats_dict and stats_dict['bdevs']:
                     stats.append( stats_dict['bdevs'][0])

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/simplyblock_core/services/new_device_discovery.py

@@ -1,25 +1,13 @@
 # coding=utf-8
-import logging
 import time
-import sys
 
-from simplyblock_core import constants, kv_store, storage_node_ops
+from simplyblock_core import constants, kv_store, storage_node_ops, utils
 from simplyblock_core.models.nvme_device import NVMeDevice
 from simplyblock_core.models.storage_node import StorageNode
 
-# Import the GELF logger
-from graypy import GELFTCPHandler
-
 from simplyblock_core.snode_client import SNodeClient
 
-# configure logging
-logger_handler = logging.StreamHandler(stream=sys.stdout)
-logger_handler.setFormatter(logging.Formatter('%(asctime)s: %(levelname)s: %(message)s'))
-gelf_handler = GELFTCPHandler('0.0.0.0', constants.GELF_PORT)
-logger = logging.getLogger()
-logger.addHandler(gelf_handler)
-logger.addHandler(logger_handler)
-logger.setLevel(logging.DEBUG)
+logger = utils.get_logger(__name__)
 
 # get DB controller
 db_store = kv_store.KVStore()

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/simplyblock_core/services/storage_node_monitor.py

@@ -182,8 +182,8 @@ def update_cluster_status(cluster_id):
 def set_node_online(node):
     if node.status != StorageNode.STATUS_ONLINE:
 
-        if not node.is_secondary_node and node.status == StorageNode.STATUS_UNREACHABLE and node.lvol:
-            if cluster.status in [Cluster.STATUS_ACTIVE, Cluster.STATUS_DEGRADED, Cluster.STATUS_SUSPENDED]:
+        if not node.is_secondary_node and node.status == StorageNode.STATUS_UNREACHABLE and node.lvols and \
+                cluster.status in [Cluster.STATUS_ACTIVE, Cluster.STATUS_DEGRADED, Cluster.STATUS_SUSPENDED]:
                 tasks_controller.add_node_to_auto_restart(node)
                 return
 
@@ -262,13 +262,17 @@ while True:
             if snode.status == StorageNode.STATUS_SCHEDULABLE and not ping_check and not node_api_check:
                 continue
 
-            # 3- check spdk_process
-            spdk_process = health_controller._check_spdk_process_up(snode.mgmt_ip)
+            spdk_process = False
+            if node_api_check:
+                # 3- check spdk_process
+                spdk_process = health_controller._check_spdk_process_up(snode.mgmt_ip)
             logger.info(f"Check: spdk process {snode.mgmt_ip}:5000 ... {spdk_process}")
 
-            # 4- check rpc
-            node_rpc_check = health_controller._check_node_rpc(
-                snode.mgmt_ip, snode.rpc_port, snode.rpc_username, snode.rpc_password)
+            node_rpc_check = False
+            if spdk_process:
+                # 4- check rpc
+                node_rpc_check = health_controller._check_node_rpc(
+                    snode.mgmt_ip, snode.rpc_port, snode.rpc_username, snode.rpc_password)
             logger.info(f"Check: node RPC {snode.mgmt_ip}:{snode.rpc_port} ... {node_rpc_check}")
 
             is_node_online = ping_check and node_api_check and spdk_process and node_rpc_check

{sbcli_dev-6.3.0 → sbcli_dev-6.3.2}/simplyblock_core/storage_node_ops.py

@@ -1255,14 +1255,11 @@ def add_node(cluster_id, node_ip, iface_name, data_nics_list,
         logger.error("Failed to set nvme options")
         return False
 
-    qpair = min(max(len(db_controller.get_storage_nodes_by_cluster_id(cluster_id)), 5) * snode.number_of_distribs, constants.QPAIR_COUNT)
-    if is_secondary_node:
-        qpair *= 2
-
+    qpair = cluster.qpair_count
     ret = rpc_client.transport_create("TCP", qpair)
     if not ret:
         logger.error(f"Failed to create transport TCP with qpair: {qpair}")
-        # return False
+        return False
 
     # 7- set jc singleton mask
     if snode.jc_singleton_mask:
@@ -1756,15 +1753,11 @@ def restart_storage_node(
         logger.error("Failed to set nvme options")
         return False
 
-    qpair = max(max(len(db_controller.get_storage_nodes_by_cluster_id(snode.cluster_id)), 5) * snode.number_of_distribs,
-                constants.QPAIR_COUNT)
-    if snode.is_secondary_node:
-        qpair *= 2
-
+    qpair = cluster.qpair_count
     ret = rpc_client.transport_create("TCP", qpair)
     if not ret:
         logger.error(f"Failed to create transport TCP with qpair: {qpair}")
-        # return False
+        return False
 
     # 7- set jc singleton mask
     if snode.jc_singleton_mask:
@@ -2971,7 +2964,7 @@ def recreate_lvstore(snode):
             lvol.health_check = True
         lvol.write_to_db(db_controller.kv_store)
 
-    time.sleep(2)
+    time.sleep(10)
 
     if sec_node and sec_node.status == StorageNode.STATUS_ONLINE:
         sec_rpc_client = RPCClient(sec_node.mgmt_ip, sec_node.rpc_port, sec_node.rpc_username, sec_node.rpc_password, timeout=3, retry=2)
@@ -3273,10 +3266,8 @@ def get_cluster_map(node_id):
         return False
 
     distribs_list = []
-    nodes = []
-
+    nodes = [snode]
     if snode.is_secondary_node:
-        nodes = [snode]
         for node in db_controller.get_storage_nodes():
             if node.secondary_node_id == snode.get_id():
                 for bdev in node.lvstore_stack:
@@ -3285,7 +3276,11 @@ def get_cluster_map(node_id):
                         distribs_list.extend(bdev["distribs_list"])
 
     else:
-        nodes = [snode, db_controller.get_storage_node_by_id(snode.secondary_node_id)]
+        if snode.secondary_node_id:
+            sec =  db_controller.get_storage_node_by_id(snode.secondary_node_id)
+            if sec:
+                nodes.append(sec)
+
         for bdev in snode.lvstore_stack:
             type = bdev['type']
             if type == "bdev_raid":