Migrating from Oracle to
summary
Step‑by‑step plan to move Oracle to YugabyteDB: map NUMBER(38) carefully (it can be a 128‑bit unsigned), convert CLOB/BLOB to text/bytea/jsonb with a 32MB limit or decompose large objects, and always specify primary key in CREATE TABLE because PKs define hash sharding. Use co‑located databases for many small tables, disable foreign keys/triggers before bulk load, split files and run parallel COPY, prefer batched multi‑row inserts, and cache sequences (example: cache=1 → ~100 r/s, cache=10000 → ~800 r/s). Consider Blitz/Replicant for parallel snapshot + CDC.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 0:32What do I hope to gain
- 2:57Partitioning
- 4:47Data Types
- 7:10Primary Key Definition
- 8:19Sort Order
- 11:30collation
- 11:50table space
- 12:12identity columns
- 15:17clob blob column types
- 16:31special indexes
- 17:58application considerations
- 19:20server side perfer
- 20:19concurrent connections
- 23:03load balancing
- 25:19retry transit transactions
- 27:31Hierarchical queries
- 28:24Recursive table
- 29:13Connect by
- 29:40Integration tips
- 32:08Replicant
related talks
30mPerformance: Tips & Tricks to Optimize the Speed of Your Mission-Critical Database
31mMake it Painless: YugabyteDB Database Migration Solutions
31mYugabyteDB Voyager: Simplify Data Migration to Accelerate Cloud Adoption
27mData Modeling: Master This Critical Step to Adopting a Distributed Database
30mTechnical Comparison of YugabyteDB with Oracle and PostgreSQL
29mData Modeling for Distributed SQL
37mDatabase Modernization at Cloud Scale: Minimizing Risk and Disruption
20mDatabase migration: Insights on Dual Read Write Strategy