sbcli-dev 17.2.9__tar.gz → 17.3.1__tar.gz

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sbcli-dev
-Version: 17.2.9
+Version: 17.3.1
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/sbcli_dev.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sbcli-dev
-Version: 17.2.9
+Version: 17.3.1
 Summary: CLI for managing SimplyBlock cluster
 Home-page: https://www.simplyblock.io/
 Author: Hamdy

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/simplyblock_core/env_var

@@ -1,5 +1,5 @@
 SIMPLY_BLOCK_COMMAND_NAME=sbcli-dev
-SIMPLY_BLOCK_VERSION=17.2.9
+SIMPLY_BLOCK_VERSION=17.3.1
 
 SIMPLY_BLOCK_DOCKER_IMAGE=public.ecr.aws/simply-block/simplyblock:main
 SIMPLY_BLOCK_SPDK_ULTRA_IMAGE=public.ecr.aws/simply-block/ultra:main-latest

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/simplyblock_core/scripts/docker-compose-swarm-monitoring.yml

@@ -67,27 +67,27 @@ services:
     networks:
       - monitoring-net
 
-  promagent:
-    image: simplyblock/promagent
-    environment:
-      ClusterID: "${CLUSTER_ID}"
-      ClusterIP: "HAProxy"
-      ClusterSecret: "${CLUSTER_SECRET}"
-    deploy:
-      mode: global
-      placement:
-        constraints: [node.role == manager]
-    networks:
-      - monitoring-net
-
-  pushgateway:
-    image: prom/pushgateway
-    deploy:
-      mode: global
-      placement:
-        constraints: [node.role == manager]
-    networks:
-      - monitoring-net
+#  promagent:
+#    image: simplyblock/promagent
+#    environment:
+#      ClusterID: "${CLUSTER_ID}"
+#      ClusterIP: "HAProxy"
+#      ClusterSecret: "${CLUSTER_SECRET}"
+#    deploy:
+#      mode: global
+#      placement:
+#        constraints: [node.role == manager]
+#    networks:
+#      - monitoring-net
+#
+#  pushgateway:
+#    image: prom/pushgateway
+#    deploy:
+#      mode: global
+#      placement:
+#        constraints: [node.role == manager]
+#    networks:
+#      - monitoring-net
 
   prometheus:
     image: prom/prometheus:v2.44.0

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/simplyblock_core/scripts/docker-compose-swarm.yml

@@ -113,7 +113,6 @@ services:
         constraints: [node.role == manager]
     volumes:
       - "/etc/foundationdb:/etc/foundationdb"
-      - "/etc/simplyblock:/etc/simplyblock"
     networks:
       - hostnet
     environment:
@@ -171,7 +170,6 @@ services:
         constraints: [node.role == manager]
     volumes:
       - "/etc/foundationdb:/etc/foundationdb"
-      - "/etc/simplyblock:/etc/simplyblock"
     networks:
       - hostnet
     environment:

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/simplyblock_core/scripts/prometheus.yml

@@ -1,17 +1,13 @@
 global:
-  scrape_interval: 15s
+  scrape_interval: 30s
   external_labels:
     monitor: 'codelab-monitor'
 
 scrape_configs:
-  - job_name: 'metricsgateway'
-    static_configs:
-      - targets: ['pushgateway:9091']
 
   - job_name: 'cluster_metrics'
     static_configs:
       - targets: ['HAProxy:80']
-
     honor_labels: true
     metrics_path: '/cluster/metrics'
 

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/simplyblock_core/services/capacity_and_stats_collector.py

@@ -2,16 +2,12 @@
 import os
 import time
 
-from simplyblock_core import constants, db_controller, utils, shell_utils
+from simplyblock_core import constants, db_controller, utils
 from simplyblock_core.models.nvme_device import NVMeDevice
 from simplyblock_core.models.storage_node import StorageNode
 from simplyblock_core.rpc_client import RPCClient
 from simplyblock_core.models.stats import DeviceStatObject, NodeStatObject, ClusterStatObject
 
-PROMETHEUS_MULTIPROC_DIR = constants.PROMETHEUS_MULTIPROC_DIR
-os.environ["PROMETHEUS_MULTIPROC_DIR"] = PROMETHEUS_MULTIPROC_DIR
-from prometheus_client import Gauge, CollectorRegistry, multiprocess
-
 logger = utils.get_logger(__name__)
 
 
@@ -88,17 +84,6 @@ def add_device_stats(cl, device, capacity_dict, stats_dict):
     stat_obj.write_to_db(db_controller.kv_store)
     last_object_record[device.get_id()] = stat_obj
 
-    ng = get_device_metrics()
-    for g in ng:
-        v = g.replace("device_", "")
-        if v in data:
-            ng[g].labels(cluster=cl.get_id(), snode=node.get_id(), device=device.get_id()).set(data[v])
-        elif v == "status_code":
-            ng[g].labels(cluster=cl.get_id(), snode=node.get_id(), device=device.get_id()).set(device.get_status_code())
-        elif v == "health_check":
-            ng[g].labels(cluster=cl.get_id(), snode=node.get_id(), device=device.get_id()).set(device.health_check)
-
-
     return stat_obj
 
 
@@ -133,16 +118,6 @@ def add_node_stats(node, records):
     stat_obj = NodeStatObject(data=data)
     stat_obj.write_to_db(db_controller.kv_store)
 
-    ng = get_snode_metrics()
-    for g in ng:
-        v = g.replace("snode_", "")
-        if v in data:
-            ng[g].labels(cluster=cl.get_id(), snode=node.get_id()).set(data[v])
-        elif v == "status_code":
-            ng[g].labels(cluster=cl.get_id(), snode=node.get_id()).set(node.get_status_code())
-        elif v == "health_check":
-            ng[g].labels(cluster=cl.get_id(), snode=node.get_id()).set(node.health_check)
-
     return stat_obj
 
 
@@ -172,84 +147,9 @@ def add_cluster_stats(cl, records):
     stat_obj = ClusterStatObject(data=data)
     stat_obj.write_to_db(db_controller.kv_store)
 
-    ng = get_cluster_metrics()
-    for g in ng:
-        v = g.replace("cluster_", "")
-        if v in data:
-            ng[g].labels(cluster=cl.get_id()).set(data[v])
-        elif v == "status_code":
-            ng[g].labels(cluster=cl.get_id()).set(cl.get_status_code())
-
     return stat_obj
 
 
-if not os.path.exists(PROMETHEUS_MULTIPROC_DIR):
-    shell_utils.run_command(f"mkdir -p {PROMETHEUS_MULTIPROC_DIR}")
-
-
-registry = CollectorRegistry()
-multiprocess.MultiProcessCollector(registry)
-
-io_stats_keys = [
-    "date",
-    "read_bytes",
-    "read_bytes_ps",
-    "read_io_ps",
-    "read_io",
-    "read_latency_ps",
-    "write_bytes",
-    "write_bytes_ps",
-    "write_io",
-    "write_io_ps",
-    "write_latency_ps",
-    "size_total",
-    "size_prov",
-    "size_used",
-    "size_free",
-    "size_util",
-    "size_prov_util",
-    "read_latency_ticks",
-    "record_duration",
-    "record_end_time",
-    "record_start_time",
-    "unmap_bytes",
-    "unmap_bytes_ps",
-    "unmap_io",
-    "unmap_io_ps",
-    "unmap_latency_ps",
-    "unmap_latency_ticks",
-    "write_latency_ticks",
-]
-
-ng = {}
-cg = {}
-dg = {}
-
-def get_device_metrics():
-    global dg
-    if not dg:
-        labels = ['cluster', "snode", "device"]
-        for k in io_stats_keys + ["status_code", "health_check"]:
-            dg["device_" + k] = Gauge("device_" + k, "device_" + k, labelnames=labels, registry=registry)
-    return dg
-
-def get_snode_metrics():
-    global ng
-    if not ng:
-        labels = ['cluster', "snode"]
-        for k in io_stats_keys + ["status_code", "health_check"]:
-            ng["snode_" + k] = Gauge("snode_" + k, "snode_" + k, labelnames=labels, registry=registry)
-    return ng
-
-def get_cluster_metrics():
-    global cg
-    if not cg:
-        labels = ['cluster']
-        for k in io_stats_keys + ["status_code"]:
-            cg["cluster_" + k] = Gauge("cluster_" + k, "cluster_" + k, labelnames=labels, registry=registry)
-    return cg
-
-
 
 # get DB controller
 db_controller = db_controller.DBController()

{sbcli_dev-17.2.9 → sbcli_dev-17.3.1}/simplyblock_core/services/lvol_stat_collector.py

@@ -2,79 +2,18 @@
 import time
 import os
 
-from simplyblock_core import constants, db_controller, utils, shell_utils
+from simplyblock_core import constants, db_controller, utils
 from simplyblock_core.controllers import lvol_events
 from simplyblock_core.models.cluster import Cluster
 from simplyblock_core.models.lvol_model import LVol
 from simplyblock_core.models.stats import LVolStatObject, PoolStatObject
 from simplyblock_core.models.storage_node import StorageNode
 from simplyblock_core.rpc_client import RPCClient
-PROMETHEUS_MULTIPROC_DIR = constants.PROMETHEUS_MULTIPROC_DIR
-os.environ["PROMETHEUS_MULTIPROC_DIR"] = PROMETHEUS_MULTIPROC_DIR
-from prometheus_client import Gauge, CollectorRegistry, multiprocess
-
-if not os.path.exists(PROMETHEUS_MULTIPROC_DIR):
-    shell_utils.run_command(f"mkdir -p {PROMETHEUS_MULTIPROC_DIR}")
-
 
 logger = utils.get_logger(__name__)
 
 last_object_record = {}
 
-registry = CollectorRegistry()
-multiprocess.MultiProcessCollector(registry)
-
-io_stats_keys = [
-    "date",
-    "read_bytes",
-    "read_bytes_ps",
-    "read_io_ps",
-    "read_io",
-    "read_latency_ps",
-    "write_bytes",
-    "write_bytes_ps",
-    "write_io",
-    "write_io_ps",
-    "write_latency_ps",
-    "size_total",
-    "size_prov",
-    "size_used",
-    "size_free",
-    "size_util",
-    "size_prov_util",
-    "read_latency_ticks",
-    "record_duration",
-    "record_end_time",
-    "record_start_time",
-    "unmap_bytes",
-    "unmap_bytes_ps",
-    "unmap_io",
-    "unmap_io_ps",
-    "unmap_latency_ps",
-    "unmap_latency_ticks",
-    "write_latency_ticks",
-]
-
-lg = {}
-pg = {}
-
-
-def get_lvol_metrics():
-    global lg
-    if not lg:
-        labels = ['cluster', "pool", "lvol"]
-        for k in io_stats_keys + ["status_code", "health_check"]:
-            lg["lvol_" + k] = Gauge("lvol_" + k, "lvol_" + k, labelnames=labels, registry=registry)
-    return lg
-
-def get_pool_metrics():
-    global pg
-    if not pg:
-        labels = ['cluster', "pool", "name"]
-        for k in io_stats_keys + ["status_code"]:
-            pg["pool_" + k] = Gauge("pool_" + k, "pool_" + k, labelnames=labels, registry=registry)
-    return pg
-
 
 def sum_stats(stats_list):
     if not stats_list or len(stats_list) == 0:
@@ -216,16 +155,6 @@ def add_lvol_stats(cluster, lvol, stats_list, capacity_dict=None):
     stat_obj.write_to_db(db_controller.kv_store)
     last_object_record[lvol.get_id()] = stat_obj
 
-    ng = get_lvol_metrics()
-    for g in ng:
-        v = g.replace("lvol_", "")
-        if v in data:
-            ng[g].labels(cluster=cluster.get_id(), lvol=lvol.get_id(), pool=lvol.pool_name).set(data[v])
-        elif v == "status_code":
-            ng[g].labels(cluster=cluster.get_id(), lvol=lvol.get_id(), pool=lvol.pool_name).set(lvol.get_status_code())
-        elif v == "health_check":
-            ng[g].labels(cluster=cluster.get_id(), lvol=lvol.get_id(), pool=lvol.pool_name).set(lvol.health_check)
-
     return stat_obj
 
 
@@ -245,15 +174,6 @@ def add_pool_stats(pool, records):
 
     stat_obj = PoolStatObject(data=data)
     stat_obj.write_to_db(db_controller.kv_store)
-
-    ng = get_pool_metrics()
-    for g in ng:
-        v = g.replace("pool_", "")
-        if v in data:
-            ng[g].labels(cluster=cluster.get_id(), name=pool.pool_name, pool=pool.get_id()).set(data[v])
-        elif v == "status_code":
-            ng[g].labels(cluster=cluster.get_id(), name=pool.pool_name, pool=pool.get_id()).set(pool.get_status_code())
-
     return stat_obj
 
 

sbcli_dev-17.3.1/simplyblock_web/blueprints/web_api_metrics.py

@@ -0,0 +1,197 @@
+#!/usr/bin/env python
+# encoding: utf-8
+import json
+import logging
+import os
+
+from flask import Blueprint
+from simplyblock_core.models.nvme_device import NVMeDevice
+from simplyblock_core.models.storage_node import StorageNode
+from simplyblock_core import db_controller
+
+from prometheus_client import generate_latest
+from flask import Response
+from prometheus_client import Gauge, CollectorRegistry
+
+
+logger = logging.getLogger(__name__)
+
+bp = Blueprint("metrics", __name__)
+
+registry = CollectorRegistry()
+db_controller = db_controller.DBController()
+
+io_stats_keys = [
+    "date",
+    "read_bytes",
+    "read_bytes_ps",
+    "read_io_ps",
+    "read_io",
+    "read_latency_ps",
+    "write_bytes",
+    "write_bytes_ps",
+    "write_io",
+    "write_io_ps",
+    "write_latency_ps",
+    "size_total",
+    "size_prov",
+    "size_used",
+    "size_free",
+    "size_util",
+    "size_prov_util",
+    "read_latency_ticks",
+    "record_duration",
+    "record_end_time",
+    "record_start_time",
+    "unmap_bytes",
+    "unmap_bytes_ps",
+    "unmap_io",
+    "unmap_io_ps",
+    "unmap_latency_ps",
+    "unmap_latency_ticks",
+    "write_latency_ticks",
+]
+
+ng = {}
+cg = {}
+dg = {}
+lg = {}
+pg = {}
+
+def get_device_metrics():
+    global dg
+    if not dg:
+        labels = ['cluster', "snode", "device"]
+        for k in io_stats_keys + ["status_code", "health_check"]:
+            dg["device_" + k] = Gauge("device_" + k, "device_" + k, labelnames=labels, registry=registry)
+    return dg
+
+def get_snode_metrics():
+    global ng
+    if not ng:
+        labels = ['cluster', "snode"]
+        for k in io_stats_keys + ["status_code", "health_check"]:
+            ng["snode_" + k] = Gauge("snode_" + k, "snode_" + k, labelnames=labels, registry=registry)
+    return ng
+
+def get_cluster_metrics():
+    global cg
+    if not cg:
+        labels = ['cluster']
+        for k in io_stats_keys + ["status_code"]:
+            cg["cluster_" + k] = Gauge("cluster_" + k, "cluster_" + k, labelnames=labels, registry=registry)
+    return cg
+
+def get_lvol_metrics():
+    global lg
+    if not lg:
+        labels = ['cluster', "pool", "lvol"]
+        for k in io_stats_keys + ["status_code", "health_check"]:
+            lg["lvol_" + k] = Gauge("lvol_" + k, "lvol_" + k, labelnames=labels, registry=registry)
+    return lg
+
+def get_pool_metrics():
+    global pg
+    if not pg:
+        labels = ['cluster', "pool", "name"]
+        for k in io_stats_keys + ["status_code"]:
+            pg["pool_" + k] = Gauge("pool_" + k, "pool_" + k, labelnames=labels, registry=registry)
+    return pg
+
+
+@bp.route('/cluster/metrics', methods=['GET'])
+def get_data():
+
+    clusters = db_controller.get_clusters()
+    for cl in clusters:
+
+        records = db_controller.get_cluster_stats(cl, 1)
+        if records:
+            data = records[0].get_clean_dict()
+            ng = get_cluster_metrics()
+            for g in ng:
+                v = g.replace("cluster_", "")
+                if v in data:
+                    ng[g].labels(cluster=cl.get_id()).set(data[v])
+                elif v == "status_code":
+                    ng[g].labels(cluster=cl.get_id()).set(cl.get_status_code())
+
+        snodes = db_controller.get_storage_nodes_by_cluster_id(cl.get_id())
+        for node in snodes:
+            logger.info("Node: %s", node.get_id())
+            if node.status != StorageNode.STATUS_ONLINE:
+                logger.info("Node is not online, skipping")
+                continue
+
+            if not node.nvme_devices:
+                logger.error("No devices found in node: %s", node.get_id())
+                continue
+
+            records = db_controller.get_node_stats(node, 1)
+            if records:
+                data = records[0].get_clean_dict()
+                ng = get_snode_metrics()
+                for g in ng:
+                    v = g.replace("snode_", "")
+                    if v in data:
+                        ng[g].labels(cluster=cl.get_id(), snode=node.get_id()).set(data[v])
+                    elif v == "status_code":
+                        ng[g].labels(cluster=cl.get_id(), snode=node.get_id()).set(node.get_status_code())
+                    elif v == "health_check":
+                        ng[g].labels(cluster=cl.get_id(), snode=node.get_id()).set(node.health_check)
+
+            for device in node.nvme_devices:
+
+                logger.info("Getting device stats: %s", device.uuid)
+                if device.status not in [NVMeDevice.STATUS_ONLINE, NVMeDevice.STATUS_READONLY, NVMeDevice.STATUS_CANNOT_ALLOCATE]:
+                    logger.info(f"Device is skipped: {device.get_id()} status: {device.status}")
+                    continue
+
+                records= db_controller.get_device_stats(device, 1)
+                if records:
+                    data = records[0].get_clean_dict()
+                    ng = get_device_metrics()
+                    for g in ng:
+                        v = g.replace("device_", "")
+                        if v in data:
+                            ng[g].labels(cluster=cl.get_id(), snode=node.get_id(), device=device.get_id()).set(data[v])
+                        elif v == "status_code":
+                            ng[g].labels(cluster=cl.get_id(), snode=node.get_id(), device=device.get_id()).set(
+                                device.get_status_code())
+                        elif v == "health_check":
+                            ng[g].labels(cluster=cl.get_id(), snode=node.get_id(), device=device.get_id()).set(
+                                device.health_check)
+
+
+        for pool in db_controller.get_pools():
+
+            records = db_controller.get_pool_stats(pool, 1)
+            if records:
+                data = records[0].get_clean_dict()
+                ng = get_pool_metrics()
+                for g in ng:
+                    v = g.replace("pool_", "")
+                    if v in data:
+                        ng[g].labels(cluster=cl.get_id(), name=pool.pool_name, pool=pool.get_id()).set(data[v])
+                    elif v == "status_code":
+                        ng[g].labels(cluster=cl.get_id(), name=pool.pool_name, pool=pool.get_id()).set(
+                            pool.get_status_code())
+
+        for lvol in db_controller.get_lvols(cl.get_id()):
+            records = db_controller.get_lvol_stats(lvol, limit=1)
+            if records:
+                data = records[0].get_clean_dict()
+                ng = get_lvol_metrics()
+                for g in ng:
+                    v = g.replace("lvol_", "")
+                    if v in data:
+                        ng[g].labels(cluster=cl.get_id(), lvol=lvol.get_id(), pool=lvol.pool_name).set(data[v])
+                    elif v == "status_code":
+                        ng[g].labels(cluster=cl.get_id(), lvol=lvol.get_id(), pool=lvol.pool_name).set(
+                            lvol.get_status_code())
+                    elif v == "health_check":
+                        ng[g].labels(cluster=cl.get_id(), lvol=lvol.get_id(), pool=lvol.pool_name).set(
+                            lvol.health_check)
+
+
+    return Response(generate_latest(registry), mimetype=str('text/plain; version=0.0.4; charset=utf-8'))

sbcli_dev-17.2.9/simplyblock_web/blueprints/web_api_metrics.py

@@ -1,31 +0,0 @@
-#!/usr/bin/env python
-# encoding: utf-8
-import json
-import logging
-import os
-
-from flask import Blueprint
-from simplyblock_core import constants, shell_utils
-
-PROMETHEUS_MULTIPROC_DIR = constants.PROMETHEUS_MULTIPROC_DIR
-os.environ["PROMETHEUS_MULTIPROC_DIR"] = PROMETHEUS_MULTIPROC_DIR
-
-from prometheus_client import generate_latest, multiprocess
-from flask import Response
-from prometheus_client import CollectorRegistry
-
-
-if not os.path.exists(PROMETHEUS_MULTIPROC_DIR):
-    shell_utils.run_command(f"mkdir -p {PROMETHEUS_MULTIPROC_DIR}")
-
-
-logger = logging.getLogger(__name__)
-
-bp = Blueprint("metrics", __name__)
-
-registry = CollectorRegistry()
-multiprocess.MultiProcessCollector(registry)
-
-@bp.route('/cluster/metrics', methods=['GET'])
-def get_data():
-    return Response(generate_latest(registry), mimetype=str('text/plain; version=0.0.4; charset=utf-8'))