BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.devconf.info//devconf-us-2026//talk//ZZBY9J
BEGIN:VTIMEZONE
TZID:EST
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10;UNTIL=20061029T070000Z
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
END:STANDARD
BEGIN:STANDARD
DTSTART:20071104T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000402T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4;UNTIL=20060402T080000Z
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20070311T030000
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-devconf-us-2026-ZZBY9J@pretalx.devconf.info
DTSTART;TZID=EST:20260925T131000
DTEND;TZID=EST:20260925T134500
DESCRIPTION:In this talk\, we will introduce vLLM and explore why monitorin
 g matters for LLM serving. We'll start from why serving LLMs is hard in th
 e first place and build up to what that complexity looks like as real\, ob
 servable metrics.\n\nvLLM is a high-performance library designed for LLM i
 nference and serving. Its core optimizations\, including PagedAttention\, 
 continuous batching\, and memory management\, directly shape the behavior 
 you'll see in production. Understanding what those techniques are doing un
 der the hood makes it easier to understand what your metrics are actually 
 telling you.\n\nThe session will include a demo showing how to instrument 
 vLLM with observability tooling and interpret the signals that matter: KV 
 cache usage\, request queue depth\, and end-to-end latency. No GPU or prod
 uction cluster required. Just a small model running locally and some pract
 ical monitoring tools.\n\nBy the end of this session\, you'll have a clear
 er mental model of LLM serving\, a working understanding of vLLM's built-i
 n observability\, and the monitoring foundation to understand what's happe
 ning inside your server before something goes wrong.
DTSTAMP:20260727T175219Z
LOCATION:Ladd Room (Capacity 170)
SUMMARY:Portrait of a Server on Fire - Rafael Vasquez
URL:https://pretalx.devconf.info/devconf-us-2026/talk/ZZBY9J/
END:VEVENT
END:VCALENDAR