← Back to Docs
Reference

Migration Settings Reference

Complete reference for every configuration option in DBMigrateAIPro. Settings are configured in the right panel of the desktop app or in the project config/settings.json file.

Connection Settings

Required

Connection settings are entered separately for source (Oracle) and target (PostgreSQL). Both connections are tested before any migration step runs.

SettingTypeDefaultDescription
hoststringrequiredHostname or IP address of the source/target database server
portint1521 / 5432Database port — 1521 for Oracle, 5432 for PostgreSQL
database / servicestringrequiredOracle service name / SID or PostgreSQL database name
usernamestringrequiredLogin username. Must have read access on Oracle, write access on PostgreSQL
passwordstringrequiredLogin password. Stored in encrypted project config, never in plaintext logs
schemastring""Source schema to migrate. Empty string migrates all accessible schemas
sslbooleanfalseEnable SSL/TLS for the database connection. Requires certificates to be configured

Performance Settings

Tuning

These settings control throughput and resource usage. Start with defaults and scale parallel_workers and batch_size based on observed rows/sec in the Perf tab.

SettingTypeDefaultDescription
batch_sizeint10000Number of rows fetched and written per batch. Lower for memory-constrained environments
parallel_workersint4Number of parallel migration threads. Valid range: 1–64. Match to CPU core count
poll_interval_secfloat5.0Seconds between progress updates in the Perf tab during live migration
memory_limit_mbint2048Soft memory ceiling in MB. Reduces batch size automatically when approached
fetchmany_sizeint5000Oracle cursor fetchmany() chunk size. Should be ≤ batch_size
bulk_methodstring"COPY" / "executemany"PostgreSQL: COPY for maximum throughput. Other targets: executemany fallback

Schema Options

Filter

Control which database objects are included in the migration. The include_tables/ exclude_tables filters are applied before DDL conversion runs.

SettingTypeDefaultDescription
include_tableslist[]Allowlist of table names to migrate. Empty list means all tables are included
exclude_tableslist[]Denylist of table names to skip. Applied after include_tables filter
include_schemastring""Migrate only objects belonging to this schema. Overrides per-connection schema setting
exclude_system_tablesbooltrueSkip Oracle system / data dictionary tables (SYS, SYSTEM schemas)
migrate_viewsbooltrueConvert and recreate Oracle views as PostgreSQL views
migrate_sequencesbooltrueRecreate Oracle sequences as PostgreSQL sequences, synced to current value
migrate_indexesbooltrueRecreate indexes after data load. Disabling speeds up load at cost of query perf
migrate_fksbooltrueRecreate foreign key constraints in PostgreSQL after data load completes
defer_fks_during_loadbooltrueDrop FKs before data load, recreate after. Required for circular FK graphs

CDC Configuration

Change Data Capture

CDC settings control the LogMiner (Oracle) or binlog (MySQL/MSSQL) streaming pipeline used for zero-downtime cutover. Supplemental logging must be enabled on Oracle before starting CDC.

SettingTypeDefaultDescription
source_typestringrequiredSource database type for CDC. One of: oracle, mysql, mssql
start_scnintautoOracle LogMiner start SCN. Defaults to SCN captured at migration start if not set
poll_intervalfloat5.0Seconds between LogMiner / binlog poll cycles
batch_sizeint1000Max change events processed per CDC poll cycle
supplemental_loggingboolrequiredOracle only — ALTER DATABASE ADD SUPPLEMENTAL LOG DATA must be enabled first
tableslist[]Fully-qualified table names (SCHEMA.TABLE) to capture CDC events for. Empty = all

Oracle Prerequisite

Enable supplemental logging before capturing the start SCN:

ALTER DATABASE ADD SUPPLEMENTAL LOG DATA;

Validation Settings

QA

Validation runs automatically after data load. Reports are written toreports/validation_report.html (or .json if configured). Enable checksum checks only when data integrity is paramount — they significantly increase validation time.

SettingTypeDefaultDescription
run_row_count_checkbooltrueCompare source vs target row counts for every migrated table after load
run_checksum_checkboolfalseRun MD5 checksum comparison per row. Slower but catches silent data corruption
checksum_columnslistPK columnsColumns included in row checksum. Defaults to primary key columns only
report_formatstring"html"Validation report output format. Accepts "html" for browser-readable or "json" for tooling