Distributed Replay Testing with Your Data, Your Way!
summary
Distributed Replay uses SQL Server 2012+ tools to capture a profiler trace, preprocess it into replay files, and synchronize execution across up to 16 clients to emulate production workloads. It validates whether changes actually affect performance — e.g. a TDE test shows TDE performance: one small read‑latency spike and CPU ~25%, proving acceptable impact. Setup requires editing XML config files, granting DCOM launch/activation and Distributed COM Users rights, opening a specific client firewall rule, restarting services, and collecting per‑client result logs.
AI-written summary · about
chapters (19)
- 0:00Intro
- 0:57Technical Assistance
- 2:54Explore Everything PASS Has to Offer
- 3:57Agenda
- 10:10Why - Limitations
- 11:52What - Administration Tool
- 13:18What - Controller
- 13:46What - Clients
- 14:17What - Target
- 14:52How
- 16:11Configuration - Preprocess
- 16:34Configuration - Controller
- 17:42Configuration - Client
- 18:25Configuration - Replay Configuration
- 29:58Collection
- 40:02Uses and Results
- 41:55Results - Read Latency
- 43:16Pass Summit 2015
- 44:13Resources
related talks
42mReplay Production Traffic Locally: Instantly Spot & Fix Breaking Changes
44mBenchmarking in the cloud
14mAn Off-The-Chain Execution Environment for Scalable Testing and Profiling...
24mDatabases beyond the tutorials
50mDebugging SQL Server
46mRecord-Replay API Test Cases & Data Mocks without writing code
23mShift-Left: A DevOps and Database Story
47mDisaster Recovery Strategies for SQL Server