Issues

Pin a DB connection for the Session's lifetime (with_connection) to survive implicit check-ins

eagerworks/staging_table#43 · opened

on Apr 21, 2026

Document thread-safety model (or add a mutex to Session)

eagerworks/staging_table#42 · opened

on Apr 21, 2026

Materialized view refresh helper (after_transfer: refresh_matview(:name))

eagerworks/staging_table#41 · opened

on Apr 21, 2026

Post-transfer hook guidance: AR callbacks are bypassed — what to do about search, caches, counter caches

eagerworks/staging_table#40 · opened

on Apr 21, 2026

Paranoid / discard gem compatibility: preserve deleted_at / discarded_at verbatim through staging

eagerworks/staging_table#39 · opened

on Apr 21, 2026

Verify Rails 7+ encrypted attributes + serialized types work end-to-end (regression tests + docs)

eagerworks/staging_table#38 · opened

on Apr 21, 2026

log_sql: true mode — emit the exact transfer SQL for debugging

eagerworks/staging_table#37 · opened

on Apr 21, 2026

OpenTelemetry span adapter for tracing alongside AS::Notifications

eagerworks/staging_table#36 · opened

on Apr 21, 2026

CLI / rake task for CSV/JSON → model imports without writing Ruby

eagerworks/staging_table#35 · opened

on Apr 21, 2026

ActiveJob-aware wrapper that solves 'temp tables die between jobs'

eagerworks/staging_table#34 · opened

on Apr 21, 2026

session.truncate to reuse a Session for successive imports without recreating the table

eagerworks/staging_table#32 · opened

on Apr 21, 2026

Around callbacks (around_insert, around_transfer) for wrapping with timing/metrics/retry

eagerworks/staging_table#31 · opened

on Apr 21, 2026

Per-batch callbacks (before_batch / after_batch) for checkpointing during large imports

eagerworks/staging_table#30 · opened

on Apr 21, 2026

TransferResult#raise_on_failure! helper

eagerworks/staging_table#29 · opened

on Apr 21, 2026

Query-as-source: stage derived data from a SELECT, not just rows/relations

eagerworks/staging_table#28 · opened

on Apr 21, 2026

Model-less staging: StagingTable.stage(table_name:, schema:) for tables without an AR model

eagerworks/staging_table#27 · opened

on Apr 21, 2026

Automatic retry with exponential backoff on deadlock / serialization failures

eagerworks/staging_table#26 · opened

on Apr 21, 2026

Postgres: SET CONSTRAINTS ALL DEFERRED inside the transfer transaction

eagerworks/staging_table#25 · opened

on Apr 21, 2026

Add returning: option to capture inserted/updated IDs from transfer

eagerworks/staging_table#24 · opened

on Apr 21, 2026

Conditional upsert: ON CONFLICT ... DO UPDATE SET ... WHERE <condition>

eagerworks/staging_table#23 · opened

on Apr 21, 2026

Create indexes on the staging table after bulk insert (opt-in)

eagerworks/staging_table#22 · opened

on Apr 21, 2026

Support UNLOGGED staging tables on Postgres (2–3× write speed, safe because staging is throwaway)

eagerworks/staging_table#21 · opened

on Apr 21, 2026

Postgres COPY FROM STDIN path for bulk insert (10–100× faster than INSERT VALUES)

eagerworks/staging_table#20 · opened

on Apr 21, 2026

Audit schema-qualified table names and Rails 7.1 composite primary keys

eagerworks/staging_table#19 · opened

on Apr 21, 2026

staging_table_name can exceed Postgres's 63-char identifier limit (and break on Oracle)

eagerworks/staging_table#18 · opened

on Apr 21, 2026

1

Navigation Menu

Repository navigation

Issues

Pin a DB connection for the Session's lifetime (with_connection) to survive implicit check-ins

Document thread-safety model (or add a mutex to Session)

Materialized view refresh helper (after_transfer: refresh_matview(:name))

Post-transfer hook guidance: AR callbacks are bypassed — what to do about search, caches, counter caches

Paranoid / discard gem compatibility: preserve deleted_at / discarded_at verbatim through staging

Verify Rails 7+ encrypted attributes + serialized types work end-to-end (regression tests + docs)

log_sql: true mode — emit the exact transfer SQL for debugging

OpenTelemetry span adapter for tracing alongside AS::Notifications

CLI / rake task for CSV/JSON → model imports without writing Ruby

ActiveJob-aware wrapper that solves 'temp tables die between jobs'

session.truncate to reuse a Session for successive imports without recreating the table

Around callbacks (around_insert, around_transfer) for wrapping with timing/metrics/retry

Per-batch callbacks (before_batch / after_batch) for checkpointing during large imports

TransferResult#raise_on_failure! helper

Query-as-source: stage derived data from a SELECT, not just rows/relations

Model-less staging: StagingTable.stage(table_name:, schema:) for tables without an AR model

Automatic retry with exponential backoff on deadlock / serialization failures

Postgres: SET CONSTRAINTS ALL DEFERRED inside the transfer transaction

Add returning: option to capture inserted/updated IDs from transfer

Conditional upsert: ON CONFLICT ... DO UPDATE SET ... WHERE <condition>

Create indexes on the staging table after bulk insert (opt-in)

Support UNLOGGED staging tables on Postgres (2–3× write speed, safe because staging is throwaway)

Postgres COPY FROM STDIN path for bulk insert (10–100× faster than INSERT VALUES)

Audit schema-qualified table names and Rails 7.1 composite primary keys

staging_table_name can exceed Postgres's 63-char identifier limit (and break on Oracle)