nvidia tested and working, added options for web and api port in jenkinsfile

README.md

@@ -6,3 +6,7 @@ The web dashboard is built from a stack of Node.js, PHP, and nginx. The Node.js
 
 The docker stack also uses a network on the 192.168.37.0/24 subnet, and secures all traffic to be within this subnet. The dashboard can only be accessed locally at port 80 on 192.168.37.1 when the secure_api variable is set to true.
 
+Working on the network descriptor, and I think to make it easier on me i will use json output from ip
+The problem is that if an interface has multiple IPs, then I have multiple lines of data for IPs
+I think perhaps I will update my code to be aware of properties that may be arrays
+I do think this makes more sense than having multiple cards for a single interface...

defaults/main.yaml

@@ -13,6 +13,7 @@ cosmostat_packages:
   - lm-sensors
   - jc
   - smartmontools
+  - inxi
 
 # python venv packages
 cosmostat_venv_packages: |
@@ -47,10 +48,12 @@ api_service_name: "cosmostat_api"
 api_service_folder: "{{ service_folder }}/api"
 venv_folder: "{{ service_folder }}/venv"
 api_service_exe: "{{ venv_folder }}/bin/python -u {{ api_service_folder }}/app.py"
+custom_api_port: "5000"
 
 # dashboard vars
 service_control_web_folder: "{{ service_folder }}/web"
 public_dashboard: true
+custom_port: "80"
 
 # will skip init when true
 quick_refresh: false

files/api/Components.py

@@ -49,16 +49,25 @@ class Component:
         # store static properties
         self.multi_check = self.is_multi()
         self.virt_ignore = self._descriptor.get('virt_ignore', [])
+        self.multi_metrics = self._descriptor.get('multi_metrics', [])
+        #if 'precheck' in self._descriptor:
+        #    precheck_command = self._descriptor.get('precheck', [])
+        #    precheck_value = int(run_command(precheck_command, zero_only = True))
+        #    if precheck_value == 0:
+        #        raise ValueError(f"No devices of type {self.type}")
         if self.is_virtual:
             self.virt_ignore = []
-        self._properties: Dict[str, str] = {}
+        self._properties: Dict[str, str | list[str]] = {}
         for key, command in descriptor.get('properties', {}).items():
+            return_string = True
+            if key in self.multi_metrics:
+                return_string = False
             if self.this_device != "None":
                 # this means this component type is a multi and the commands need templating for each device
                 formatted_command = command.format(this_device=self.this_device)
-                self._properties[key] = run_command(formatted_command, True)       
+                self._properties[key] = run_command(formatted_command, zero_only = return_string)
             else:
-                self._properties[key] = run_command(command, zero_only = True)
+                self._properties[key] = run_command(command, zero_only = return_string)
             print(self._properties[key])
         # build the description string
         self._description_template: str | None = descriptor.get("description")
@@ -114,31 +123,32 @@ class Component:
             component_properties = self._properties.items()
         else:
             component_properties = self.get_property(component)
-        for name, value in component_properties:
+        for name, values in component_properties:
-            this_property = {
+            for value in (values if isinstance(values, list) else [values]):
-                "Source": self.name,
+                this_property = {
-                "Property": name,
+                    "Source": self.name,
-                "Value": value
+                    "Property": name,
-            }  
+                    "Value": value
-            if name not in self.virt_ignore: 
+                }  
-                result.append(this_property)
+                if name not in self.virt_ignore: 
+                    result.append(this_property)
         return result
 
     def get_properties_strings(self, return_simple = False):
         result = []
         component_properties = self._properties.items()
-        print(component_properties)
+        for name, values in component_properties:
-        for name, value in component_properties:
+            for value in (values if isinstance(values, list) else [values]):
-            simple_property = f"{name}: {value}"
+                simple_property = f"{name}: {value}"
-            complex_property = {
+                complex_property = {
-                "Source": self.name,
+                    "Source": self.name,
-                "Property": simple_property
+                    "Property": simple_property
-            }  
+                }  
-            if name not in self.virt_ignore:
+                if name not in self.virt_ignore:
-                if return_simple:
+                    if return_simple:
-                    result.append(simple_property)
+                        result.append(simple_property)
-                else:
+                    else:
-                    result.append(complex_property)
+                        result.append(complex_property)
         return result
 
     def get_metrics_keys(self):
@@ -318,15 +328,15 @@ class System:
             multi_check = component["multi_check"]
             # if multi, note that the command in device_list creates the list of things to pipe into this_device
             if multi_check:
-                letters = [chr(c) for c in range(ord('A'), ord('Z')+1)]
                 print(f"Creating one component of type {component_name} for each one found")
                 component_type_device_list = get_device_list(component_name)
-
+                component_id = 0
                 for this_device in component_type_device_list:
-                    this_component_letter = letters[component_type_device_list.index(this_device)]
+                    this_component_ID = component_type_device_list.index(this_device)
-                    this_component_name = f"{component_name} {this_component_letter}"
+                    this_component_name = f"{component_name} {this_component_ID}"
                     print(f"{this_component_name} - {component_name} - {this_device}")
-                    self.add_components(Component(name = this_component_name, comp_type = component_name, this_device = this_device))
+                    new_component = Component(name = this_component_name, comp_type = component_name, this_device = this_device)
+                    self.add_components(new_component)
 
             else:
                 if debug_output:
@@ -538,7 +548,13 @@ def run_command(cmd, zero_only=False, use_shell=True, req_check = True):
 def get_device_list(device_type_name: str):
     result = []
     for component in component_class_tree:
-        if component["name"] == device_type_name:
+        precheck_value = 1
+        if "precheck" in component:
+            precheck_command = component["precheck"]
+            precheck_value_output = run_command(precheck_command, zero_only = True)
+            precheck_value = int(precheck_value_output)
+            print(f"Precheck found - {precheck_command} - {precheck_value}")        
+        if component["name"] == device_type_name and precheck_value != 0:
             device_list_command = component["device_list"]
             device_list_result = run_command(device_list_command)
             result = device_list_result

files/api/app.py

@@ -21,7 +21,8 @@ app_settings = {
     "secure_api" : True,
     "push_redis" : False,
     "run_background" : True,
-    "update_frequency": 1
+    "update_frequency": 1,
+    "custom_api_port": "5000"
     }
 
 with open('cosmostat_settings.yaml', 'r') as f:
@@ -60,6 +61,9 @@ def service_gateway_ip():
     else:
         return "0.0.0.0"
 
+def service_api_port():
+    return cosmostat_settings["custom_api_port"]
+
 #######################################################################
 ### Redis Functions
 #######################################################################
@@ -295,7 +299,7 @@ if __name__ == '__main__':
             print("Skipping flask background task")
 
     # Flask API
-    app.run(debug=False, host=service_gateway_ip(), port=5000)
+    app.run(debug=False, host=service_gateway_ip(), port=service_api_port())
 
     
 

files/api/component_descriptors.json

@@ -51,5 +51,48 @@
     "metrics": {
       "placeholder": ""
     }
+  },
+  {
+    "name": "LAN",
+    "description": "{Device ID} - {Device Name} - {MAC Address}",
+    "multi_check": "True",
+    "device_list": "ip link | grep default | grep -v -e docker -e 127.0.0.1 -e br- -e veth -e lo -e tun | cut -d ':' -f 2 | awk '{{print $1}}' ",
+    "properties": {
+      "MAC Address": "ip link | grep -A1 ' {this_device}' | grep ether | awk '{{print $2}}'",
+      "Device Name": "echo {this_device}",
+      "Device ID": "udevadm info -q property -p  $(ls -l /sys/class/net/ | grep {this_device} | cut -d '>' -f2 | cut -b 8- ) | grep ID_MODEL_FROM_DATABASE | cut -d '=' -f2 "
+    },
+    "metrics": {
+      "IP Address": "ip -o -4 ad | grep -v -e docker -e 127.0.0.1 -e br- | grep {this_device} | awk '{{print $4}}'",
+      "Data Transmitted": "ifconfig {this_device} | grep RX | grep bytes | cut -d '(' -f2 | tr -d ')'",
+      "Data Received": "ifconfig {this_device} | grep TX | grep bytes | cut -d '(' -f2 | tr -d ')'",
+      "Link State": "cat /sys/class/net/{this_device}/operstate",
+      "Link Speed": "cat /sys/class/net/{this_device}/speed"
+    },  
+    "multi_metrics": [
+        "IP Address"
+      ]
+  },  
+  {
+    "name": "NVGPU",
+    "description": "NVGPU{Device ID} - {Device Model} with {Memory Size}, Max Power {Maximum Power}",
+    "multi_check": "True",
+    "device_list": "nvidia-smi --query-gpu=index --format=csv,noheader,nounits",
+    "properties": {
+      "Device Model": "nvidia-smi --id={this_device} --query-gpu=name --format=csv,noheader,nounits",
+      "Device ID": "echo NVGPU{this_device}",
+      "Driver Version": "nvidia-smi --id={this_device} --query-gpu=driver_version --format=csv,noheader,nounits",
+      "Maximum Power": "nvidia-smi --id={this_device} --query-gpu=power.draw --format=csv,noheader,nounits",
+      "Memory Size": "nvidia-smi --id={this_device} --query-gpu=memory.total --format=csv,noheader,nounits"
+
+    },
+    "metrics": {
+      "Power Draw": "nvidia-smi --id={this_device} --query-gpu=power.draw --format=csv,noheader,nounits",
+      "Used Memory": "nvidia-smi --id={this_device} --query-gpu=memory.used --format=csv,noheader,nounits",
+      "Temperature": "nvidia-smi --id={this_device} --query-gpu=temperature.gpu --format=csv,noheader,nounits",
+      "GPU Load": "nvidia-smi --id={this_device} --query-gpu=utilization.gpu --format=csv,noheader,nounits"
+
+    },
+    "precheck": "lspci | grep NV | wc -l"
   }
 ]

files/api/new_descriptors.json

@@ -1,24 +1,22 @@
 [
   {
-    "name": "LAN",
+    "name": "",
     "description": "",
     "multi_check": "True",
-    "device_list": "",
+    "device_list": " ",
     "properties": {
-      "MAC Address": "",
+
-      "Device Name": "",
-      "Device ID": ""
     },
     "metrics": {
-      "IP Address": "",
+    },  
-      "MB Transmitted": "",
+    "multi_metrics": [
-      "MB Received": "",
+      ],
-      "Link State": "",
+    "virt_ignore": [      
-      "Link Speed": ""
+    ]
-    }
   },
   {
       "SATA GBW": "sudo /usr/sbin/smartctl -x --json /dev/{this_device} | jq -r '.physical_block_size as $block |.ata_device_statistics.pages[] | select(.name == \"General Statistics\") | .table[] | select(.name == \"Logical Sectors Written\") | .value as $sectors | ($sectors * $block) / 1073741824 ' | awk '{{printf \"%.2f GiB Written\\n\", $0}}' || true",
       "NVMe GBW": "sudo /usr/sbin/smartctl -x --json /dev/{this_device} | jq -r ' .nvme_smart_health_information_log.data_units_written as $dw | .logical_block_size as $ls | ($dw * $ls) / 1073741824  ' | awk '{{printf \"%.2f GiB Written\\n\", $0}}' || true"
   }
-]
+]
+

files/web/html/index.php

@@ -2,7 +2,7 @@
 <html lang="en">
 <head>
   <meta charset="utf-8">
-  <title>Matt-Cloud Cosmostat</title>
+  <title>Cosmostat - <?php echo $_SERVER['SERVER_NAME'] ?></title>
 
   <style>
       .components {display:grid; grid-template-columns:repeat(auto-fill, minmax(280px, 1fr)); gap:1rem;}
@@ -27,8 +27,28 @@
       <!-- PHP to render static components -->
 
       <?php
+      # load API settings, this requires a simple yaml file
+      $raw_api_settings = file('/opt/api_settings/cosmostat_settings.yaml', FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES);
+      $api_settings = [];
+      foreach ($raw_api_settings as $line) {
+          if ($line[0] === '#') {
+              continue;
+          }
+          $pos = strpos($line, ':');
+          if ($pos === false) {
+              continue;
+          }
+          $key   = trim(substr($line, 0, $pos));
+          $value = trim(substr($line, $pos + 1));
+          if ($value === '') {
+              $value = null;
+          }
+          $api_settings[$key] = $value;
+      }
+      $dockerGateway = trim($api_settings['docker_gateway'], "\"'") ?? null;
+      $customApiPort = trim($api_settings['custom_api_port'], "\"'") ?? null;
       # load API data
-      $apiUrl  = 'http://192.168.37.1:5000/php_summary';
+      $apiUrl  = 'http://'.$dockerGateway.':'.$customApiPort.'/php_summary';
       $context = stream_context_create([
           'http' => [
               'timeout' => 5,      // seconds

files/web/node_server/package.json

@@ -9,6 +9,7 @@
     "express": "^4.18.2",
     "socket.io": "^4.7.2",
     "redis": "^4.6.7",
-    "node-fetch": "^2.6.7" 
+    "node-fetch": "^2.6.7",
+    "js-yaml": "^4.1.0" 
   }
 }

files/web/node_server/server.js

@@ -4,21 +4,39 @@ const express = require('express');
 const { createClient } = require('redis');
 const { Server } = require('socket.io');
 const fetch  = require('node-fetch');   // npm i node-fetch@2
+const fs     = require('fs');
+const yaml   = require('js-yaml');      // npm i js-yaml
+const path   = require('path');
 
 const app    = express();
 const server = http.createServer(app);
 const io     = new Server(server);
 
-// ---------- Socket.io ----------
+/* --------------------------------------------------------------------- */
+/* ---------- 1. Load the YAML configuration file ---------------------- */
+/* --------------------------------------------------------------------- */
+let config = {};
+try {
+  const file = fs.readFileSync(path.resolve(__dirname, 'cosmostat_settings.yaml'), 'utf8');
+  config = yaml.load(file);
+} catch (e) {
+  console.error('Failed to load config.yaml:', e);
+  process.exit(1);
+}
+
+const API_PORT   = config.custom_api_port   || 5000;   // fallback to 5000
+const API_HOST   = config.docker_gateway   || '192.168.37.1'; // fallback IP
+const API_BASE   = `http://${API_HOST}:${API_PORT}`;
+
+// ---------------------------------------------------------------------
+// ---------- 2. Socket.io ------------------------------------------------
+// ---------------------------------------------------------------------
 io.on('connection', async socket => {
   console.log('client connected:', socket.id);
 
   // Call the external API every time a client connects
   try {
-    const resp = await fetch('http://192.168.37.1:5000/start_timer', {
+    const resp = await fetch(`${API_BASE}/start_timer`, { method: 'GET' });
-      method: 'GET'
-    });
-
     const data = await resp.json();
     console.log('API responded to connect:', data);
   } catch (err) {
@@ -29,9 +47,7 @@ io.on('connection', async socket => {
   socket.on('tableRendered', async () => {
     console.log('Client reported table rendered - starting timer');
     try {
-      const resp = await fetch('http://192.168.37.1:5000/start_timer', {
+      const resp = await fetch(`${API_BASE}/start_timer`, { method: 'GET' });
-        method: 'GET'
-      });
       const text = await resp.text();
       console.log('Timer endpoint responded:', text);
     } catch (err) {
@@ -40,17 +56,19 @@ io.on('connection', async socket => {
   });
 });
 
-// Serve static files (index.html, etc.)
+/* --------------------------------------------------------------------- */
+/* ---------- 3. Serve static files ----------------------------------- */
+/* --------------------------------------------------------------------- */
 app.use(express.static('public'));
 
-// ---------- Redis subscriber ----------
+/* --------------------------------------------------------------------- */
+/* ---------- 4. Redis subscriber ------------------------------------- */
+/* --------------------------------------------------------------------- */
 const redisClient = createClient({ url: 'redis://192.168.37.1:6379' });
-
 redisClient.on('error', err => console.error('Redis error', err));
 
 (async () => {
   await redisClient.connect();
-
   const sub = redisClient.duplicate(); // duplicate to keep separate pub/sub
   await sub.connect();
 
@@ -72,7 +90,9 @@ redisClient.on('error', err => console.error('Redis error', err));
   sub.on('error', err => console.error('Subscriber error', err));
 })();
 
-// ---------- Start ----------
+/* --------------------------------------------------------------------- */
+/* ---------- 5. Start the HTTP server --------------------------------- */
+/* --------------------------------------------------------------------- */
 const PORT = process.env.PORT || 3000;
 server.listen(PORT, () => {
   console.log(`Server listening on http://localhost:${PORT}`);

templates/cosmostat_settings.yaml

@@ -29,4 +29,5 @@ push_redis: {{ push_redis }}
 run_background : {{ run_background }}
 log_output: {{ log_output }}
 update_frequency: {{ update_frequency }}
+custom_api_port: {{ custom_api_port }}
 ...

templates/docker-compose-php.yaml

@@ -17,6 +17,7 @@ services:
     volumes:
       - "{{ service_control_web_folder }}/html:/usr/src/app/public"
       - "{{ service_control_web_folder }}/node_server:/app"
+      - "{{ api_service_folder }}/cosmostat_settings.yaml:/app/cosmostat_settings.yaml:ro"
       - /app/node_modules
     ports:
       - "{{ docker_gateway }}:3000:3000"
@@ -33,6 +34,7 @@ services:
       - "{{ docker_gateway }}:8080:80"
     volumes:
       - ./html:/var/www/html/
+      - "{{ api_service_folder }}/cosmostat_settings.yaml:/opt/api_settings/cosmostat_settings.yaml:ro"
     networks:
       - cosmostat_net
     restart: always    
@@ -42,7 +44,7 @@ services:
     container_name: cosmostat_nginx_proxy
     image: nginx:latest
     ports:
-      - "{{ (docker_gateway + ':') if not public_dashboard | bool else '' }}80:80"
+      - "{{ (docker_gateway + ':') if not public_dashboard | bool else '' }}{{ custom_port }}:80"
     volumes:
       - ./proxy/nginx.conf:/etc/nginx/conf.d/default.conf
     networks: