BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//sf26-agenda//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:SESS-136@sf26-agenda
DTSTAMP:20261020T151709Z
DTSTART:20261105T001500Z
DTEND:20261105T010000Z
SUMMARY:Queues\, Ordering\, and Long-Running Tasks in Kafka with Atleon
LOCATION:Room 2004
DESCRIPTION:Sage Pierce\n\nAt Indeed\, we help people get jobs. Our team su
 pports this mission by ingesting data about employers’ hiring needs\, en
 abling them to manage this data\, and transforming it into searchable job 
 advertisements—commonly known as job posts. The many different asynchron
 ous processes involved in solving these and related challenges can be comp
 lex\, widely diverse\, and in many cases\, I/O dependent. Consequently\, o
 ur team implements a wide swath of asynchronous use cases on top of Kafka.
  This includes conventional idiomatic use cases\, but just as important\, 
 use cases that are historically tricky on top of Kafka’s architecture. S
 uch use cases involve “treating Kafka like a queue” and implementing l
 ong-running (i.e. minutes\, hours) per-record consumption tasks. This talk
  will dive into how we at Indeed tackle these challenges with Kafka and At
 leon\, an OSS project\, as well as explore the following topics: The diffi
 culties involved with “treating Kafka like a queue” How “share group
 s”\, introduced by KIP-932\, may (or may not) address needs for higher c
 onsumer concurrency What reliable queue-like consumption without loss of o
 rdering looks like with vanilla Kafka consumer groups Facilitating long-ru
 nning per-record consumption tasks Accomplishing per-key processing order 
 and concurrency
URL:https://sf26.rmoff.net/s/queues-ordering-and-long-running-tasks-in-kafk
 a-with-atleon-136
END:VEVENT
END:VCALENDAR