feat(cluster,api,models,cli): add cluster-aware model catalog deployments and request routing

This commit is contained in:
2026-04-20 23:00:50 +00:00
parent 83cacd0cf1
commit 4f2266e1b7
55 changed files with 3970 additions and 1630 deletions
+62 -12
View File
@@ -25,6 +25,26 @@ export class ConfigHandler {
const configPath = PATHS.CONFIG_FILE;
const configContent = await fs.readFile(configPath, 'utf-8');
const config = JSON.parse(configContent) as IModelGridConfig;
const modelConfig = {
registryUrl: config.models.registryUrl ||
(config.models as { greenlistUrl?: string }).greenlistUrl ||
'https://list.modelgrid.com/catalog/models.json',
autoDeploy: config.models.autoDeploy ??
(config.models as { autoPull?: boolean }).autoPull ?? true,
defaultEngine: config.models.defaultEngine || 'vllm',
autoLoad: config.models.autoLoad || [],
};
const clusterConfig = config.cluster || {
enabled: false,
nodeName: 'modelgrid-local',
role: 'standalone',
bindHost: '0.0.0.0',
gossipPort: 7946,
sharedSecret: undefined,
advertiseUrl: undefined,
controlPlaneUrl: undefined,
heartbeatIntervalMs: 5000,
};
// Overview
logger.logBox(
@@ -48,9 +68,7 @@ export class ConfigHandler {
`Host: ${theme.info(config.api.host)}`,
`Port: ${theme.highlight(String(config.api.port))}`,
`API Keys: ${config.api.apiKeys.length} configured`,
...(config.api.rateLimit
? [`Rate Limit: ${config.api.rateLimit} req/min`]
: []),
...(config.api.rateLimit ? [`Rate Limit: ${config.api.rateLimit} req/min`] : []),
'',
theme.dim('Endpoint:'),
` http://${config.api.host}:${config.api.port}/v1/chat/completions`,
@@ -88,12 +106,33 @@ export class ConfigHandler {
logger.logBox(
'Models',
[
`Auto Pull: ${config.models.autoPull ? theme.success('Enabled') : theme.dim('Disabled')}`,
`Default Container: ${config.models.defaultContainer}`,
`Auto Load: ${config.models.autoLoad.length} model(s)`,
`Auto Deploy: ${
modelConfig.autoDeploy ? theme.success('Enabled') : theme.dim('Disabled')
}`,
`Default Engine: ${modelConfig.defaultEngine}`,
`Auto Load: ${modelConfig.autoLoad.length} model(s)`,
'',
theme.dim('Greenlist URL:'),
` ${config.models.greenlistUrl}`,
theme.dim('Registry URL:'),
` ${modelConfig.registryUrl}`,
],
70,
'default',
);
logger.log('');
logger.logBox(
'Cluster',
[
`Enabled: ${clusterConfig.enabled ? theme.success('Yes') : theme.dim('No')}`,
`Node: ${clusterConfig.nodeName}`,
`Role: ${clusterConfig.role}`,
`Bind Host: ${clusterConfig.bindHost}:${clusterConfig.gossipPort}`,
`Shared Secret: ${
clusterConfig.sharedSecret ? theme.success('Configured') : theme.dim('Not set')
}`,
`Advertise URL: ${clusterConfig.advertiseUrl || theme.dim('Default loopback')}`,
`Control Plane: ${clusterConfig.controlPlaneUrl || theme.dim('Not configured')}`,
`Heartbeat: ${clusterConfig.heartbeatIntervalMs}ms`,
],
70,
'default',
@@ -110,7 +149,7 @@ export class ConfigHandler {
name: c.name,
type: c.type,
image: c.image.length > 40 ? c.image.substring(0, 37) + '...' : c.image,
port: c.port,
port: String(c.port),
gpus: c.gpuIds.length > 0 ? c.gpuIds.join(',') : theme.dim('None'),
}));
@@ -189,11 +228,22 @@ export class ConfigHandler {
},
containers: [],
models: {
greenlistUrl: 'https://code.foss.global/modelgrid.com/model_lists/raw/branch/main/greenlit.json',
autoPull: true,
defaultContainer: 'ollama',
registryUrl: 'https://list.modelgrid.com/catalog/models.json',
autoDeploy: true,
defaultEngine: 'vllm',
autoLoad: [],
},
cluster: {
enabled: false,
nodeName: 'modelgrid-local',
role: 'standalone',
bindHost: '0.0.0.0',
gossipPort: 7946,
sharedSecret: '',
advertiseUrl: 'http://127.0.0.1:8080',
heartbeatIntervalMs: 5000,
seedNodes: [],
},
checkInterval: 30000,
};