Aerospike Management APIs and Operations at Scale
summary
PayPal automates Aerospike operations with Ansible playbooks (about 50 playbooks and 78 roles) to do rolling OS patches and DB upgrades across ~106 clusters and 1,800–2,000 hosts, reducing hands-on effort to roughly 0.5 FTE per task and enabling TLS certificate rotation and rolling software upgrades. A homegrown monitoring system uses manticore agents per cluster, feeds Grafana and alerting, and can trigger automated failover and report capacity/set-level storage. Outcome: patched 1800–2000 servers in under three months and routine quarterly cycles.
AI-written summary · about
chapters (27)
- 0:00Introduction
- 11:40Ansible Engine
- 19:05Template
- 21:20Configuration
- 23:28Additional Features
- 28:25Welcome
- 28:42Why do we need monitoring
- 30:03How does monitoring work
- 31:36Manticore
- 33:12System Report
- 33:54System Stats
- 34:27Latency Stats
- 35:56Histogram
- 36:58Capacity Report
- 37:21Dashboard
- 38:15Poll
- 38:38QA Session
- 38:58QA Question 1
- 39:54QA Question 2
- 40:36QA Question 3
- 41:56QA Question 4
- 43:31QA Question 6
- 45:09QA Question 7
- 46:03QA Question 8
- 46:30QA Question 9
- 46:49QA Question 10
- 47:47Wrapup
related talks
25mHybrid Memory Aerospike @ PayPal
05mHighlights From Aerospike APJ Summit 2023
51mLarge-scale fault-tolerant on prem deployments powering 100M+ transactions a day
49mJourney with Aerospike: Architecture Best Practices
36maccelerating the digital agenda
40mAerospike Demo Showcase
06mSnapdeal and Aerospike
29mFlipkart: Scaling Aerospike at Flipkart with Kubernetes