Are you implying that there's only one concert being served, I don't think that would be the case. I'd prefer capacity of 20k concurrent users, not 200.

It's fairly well reported that Ticketmaster's queue is based around Kafka/KSQL/Kafka Streams. Having spent since 2015 working with Kafka I know tuning this is far from a trivial task from producers, consumers and everything in between (Zookeeper, partitions and network throughput is a whole post in itself). Once you introduce mutli-rack, multi-data center complexity then it's a support task that would test even the most seasoned of staff.

An interesting post, nicely written.