Service class for tracking and logging performance metrics of client sessions in the swarm system. Monitors execution times, input/output lengths, and session states, aggregating data into IPerformanceRecord and IClientPerfomanceRecord structures. Integrates with ClientAgent workflows (e.g., execute, run) to measure performance, using LoggerService for logging (gated by GLOBAL_CONFIG.CC_LOGGER_ENABLE_INFO) and validation/public services for state computation. Provides methods to start/end executions, retrieve metrics, and serialize performance data for reporting or analytics.
constructor();
loggerService: any
Logger service instance for logging performance-related information, injected via DI. Controlled by GLOBAL_CONFIG.CC_LOGGER_ENABLE_INFO, used across methods (e.g., startExecution, toRecord) for info-level logging.
sessionValidationService: any
Session validation service instance, injected via DI. Used to retrieve session lists (e.g., getActiveSessions) and swarm names (e.g., computeClientState).
memorySchemaService: any
Memory schema service instance, injected via DI. Provides session memory data for toClientRecord, aligning with IClientPerfomanceRecord.sessionMemory.
swarmValidationService: any
Swarm validation service instance, injected via DI. Retrieves agent and policy lists for computeClientState, supporting swarm-level state aggregation.
agentValidationService: any
Agent validation service instance, injected via DI. Fetches state lists for agents in computeClientState, enabling client state computation.
statePublicService: any
State public service instance, injected via DI. Retrieves state values for computeClientState, populating IClientPerfomanceRecord.sessionState.
swarmPublicService: any
Swarm public service instance, injected via DI. Provides agent names for computeClientState, supporting swarm status in sessionState.
policyPublicService: any
Policy public service instance, injected via DI. Checks for bans in computeClientState, contributing to policyBans in sessionState.
stateConnectionService: any
State connection service instance, injected via DI. Verifies state references in computeClientState, ensuring valid state retrieval.
executionScheduleMap: any
Map tracking execution start times for clients, keyed by clientId and executionId. Used in startExecution and endExecution to calculate response times per execution.
executionOutputLenMap: any
Map of total output lengths per client, keyed by clientId. Updated in endExecution, used for IClientPerfomanceRecord.executionOutputTotal.
executionInputLenMap: any
Map of total input lengths per client, keyed by clientId. Updated in startExecution, used for IClientPerfomanceRecord.executionInputTotal.
executionCountMap: any
Map of execution counts per client, keyed by clientId. Updated in startExecution, used for IClientPerfomanceRecord.executionCount.
executionTimeMap: any
Map of total execution times per client, keyed by clientId. Updated in endExecution, used for IClientPerfomanceRecord.executionTimeTotal.
totalResponseTime: any
Total response time across all executions, in milliseconds. Aggregated in endExecution, used for IPerformanceRecord.totalResponseTime.
totalRequestCount: any
Total number of execution requests across all clients. Incremented in endExecution, used for IPerformanceRecord.totalExecutionCount.
computeClientState: any
Computes the aggregated state of a client by collecting swarm, agent, policy, and state data. Used in toClientRecord to populate IClientPerfomanceRecord.sessionState, integrating with validation and public services. Logs via loggerService if GLOBAL_CONFIG.CC_LOGGER_ENABLE_INFO is true (e.g., ClientAgent-style debug logging).
getActiveSessionExecutionCount: (clientId: string) => number
Retrieves the number of active executions for a client’s session. Used to monitor execution frequency, reflecting IClientPerfomanceRecord.executionCount.
getActiveSessionExecutionTotalTime: (clientId: string) => number
Retrieves the total execution time for a client’s sessions, in milliseconds. Used for performance analysis, feeding into IClientPerfomanceRecord.executionTimeTotal.
getActiveSessionExecutionAverageTime: (clientId: string) => number
Calculates the average execution time per execution for a client’s sessions, in milliseconds. Used for performance metrics, contributing to IClientPerfomanceRecord.executionTimeAverage.
getActiveSessionAverageInputLength: (clientId: string) => number
Calculates the average input length per execution for a client’s sessions. Used for data throughput analysis, feeding into IClientPerfomanceRecord.executionInputAverage.
getActiveSessionAverageOutputLength: (clientId: string) => number
Calculates the average output length per execution for a client’s sessions. Used for data throughput analysis, feeding into IClientPerfomanceRecord.executionOutputAverage.
getActiveSessionTotalInputLength: (clientId: string) => number
Retrieves the total input length for a client’s sessions. Used for data volume tracking, aligning with IClientPerfomanceRecord.executionInputTotal.
getActiveSessionTotalOutputLength: (clientId: string) => number
Retrieves the total output length for a client’s sessions. Used for data volume tracking, aligning with IClientPerfomanceRecord.executionOutputTotal.
getActiveSessions: () => string[]
Retrieves the list of active session client IDs. Sources data from sessionValidationService, used in toRecord to enumerate clients.
getAverageResponseTime: () => number
Calculates the average response time across all executions, in milliseconds. Used for system-wide performance metrics, feeding into IPerformanceRecord.averageResponseTime.
getTotalExecutionCount: () => number
Retrieves the total number of executions across all clients. Used for system-wide metrics, aligning with IPerformanceRecord.totalExecutionCount.
getTotalResponseTime: () => number
Retrieves the total response time across all executions, in milliseconds. Used for system-wide metrics, feeding into IPerformanceRecord.totalResponseTime.
startExecution: (executionId: string, clientId: string, inputLen: number) => void
Starts tracking an execution for a client, recording start time and input length. Initializes maps and increments execution count/input length, used with endExecution to measure performance (e.g., ClientAgent.execute).
endExecution: (executionId: string, clientId: string, outputLen: number) => boolean
Ends tracking an execution for a client, calculating response time and updating output length. Pairs with startExecution to compute execution duration, updating totals for IClientPerfomanceRecord metrics.
toClientRecord: (clientId: string) => Promise<IClientPerfomanceRecord>
Serializes performance metrics for a specific client into an IClientPerfomanceRecord. Aggregates execution counts, input/output lengths, times, memory, and state, used in toRecord for per-client data.
toRecord: () => Promise<IPerformanceRecord>
Serializes performance metrics for all clients into an IPerformanceRecord. Aggregates client records, total execution counts, and response times, used for system-wide performance reporting.
dispose: (clientId: string) => void
Disposes of all performance data associated with a client. Clears maps for the clientId, used to reset or terminate tracking (e.g., session end in ClientAgent).