Scraping EventsiTotal scraping pipeline events across all stages. Success Ratei(SUCCESS + SKIP) / Total events. Avg DurationiAverage duration of LLM_EXTRACT + RESULT_PROCESS for successful events. Active UsersiUnique users who triggered scraping tasks. Scraping ErrorsiEvents with ERROR or FAIL status. LLM CallsiTotal LLM API calls. Total TokensiSum of user_prompt_token + response_token. Cost ~$3/1M tokens. Avg LLM LatencyiAverage response_end_latency across all LLM calls. Credits ConsumediTotal credits deducted across all users. App ErrorsiTotal error count from error-log. Scraping VolumeiHourly count of all scraping pipeline events across all stages.
LLM CallsiHourly count of LLM API calls from token-metrics logstore.
Error TrendiHourly count of application errors from the error-log logstore.
Success Rate %iHourly pipeline success rate. (SUCCESS + SKIP) / Total * 100.
Stage DistributioniDistribution of successful events across pipeline stages (TASK_SUBMIT, HTML_FETCH, MD_CONVERT, LLM_EXTRACT, RESULT_PROCESS).
Model DistributioniDistribution of LLM calls across different models. Shows the top 8 most used models.