Maximizing Aerospike Administration in GCP
summary
Wayfair runs Aerospike at 60 terabytes and reports 20 million reads/sec on a fulfillment cluster, moving 60 billion records to GCP and splitting clusters by workload to avoid noisy neighbors. Automation uses Terraform + Puppet for instance and config builds, Backstage (with a Jazz layer) for self-serve user creation and replication tickets, and Datadog for metrics plus a custom Datadog user-quota import hack to show per-user reads/writes; SSDs replace all-memory as data grows.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:43Employees
- 1:04Ad Tech
- 1:57Customer Identity
- 3:48Cluster Setup
- 6:47OS
- 7:02Monitoring
- 7:52Developer Dashboard
- 8:38User Quotas
- 9:40Terraform
- 10:06Puppet
- 10:35Console
- 11:22Backstage
- 13:14Page of Duty
related talks
24mCloud Workloads At Petabyte Scale
15mAWS On Air Voice of the Customer ft. Aerospike, Inc.
09mInnovating Together for Real-time Impact with Google Cloud
49mBuilding real-time applications to utilize AWS Graviton
32mReal time Bidding Caching Use Case On The Google Cloud Platform
30mBuilding distributed systems is hard - case study (...)
33mUsing Aerospike as a Real-time System-of-Record for Adobe Experience Cloud Services on Azure
25mScaling Data Platforms: Managing Growth from Zero to Billions of Transactions