🪢 Open source agent evals & observability: Trace, evaluate, and improve LLM applications with one open platform.
TypeScript analyticsautogenevaluationlangchainlarge-language-modelsllama-indexllmllm-evaluationllm-observabilityllmopsmonitoringobservabilityopen-sourceopenaiplaygroundprompt-engineeringprompt-managementself-hostedycombinator
import { logger, sleep } from "@langfuse/shared/src/server";
/**
* Returns a uniformly distributed delay in [0, maxMs] milliseconds.
*
* Used to spread out the Redis connection setup of workers that boot at the
* same time, e.g. when autoscaling adds many tasks in one step.
*/
export const getStartupJitterMs = (
maxMs: number,
random: () => number = Math.random,
): number => {
if (maxMs <= 0) return 0;
return Math.floor(random() * (maxMs + 1));
};
/**
* Sleeps for a random startup jitter of up to `maxMs` before queue consumers
* and periodic runners open their Redis connections.
*/
export const delayStartupByJitter = async (maxMs: number): Promise<void> => {
const jitterMs = getStartupJitterMs(maxMs);
if (jitterMs <= 0) return;
logger.info(`Delaying queue registration by ${jitterMs}ms (startup jitter)`);
await sleep(jitterMs);
};