---
title: "Convert MySQL to PostgreSQL"
description: "Convert a MySQL or MariaDB dump into a PostgreSQL script that loads on the first try, with a report of every change. In your browser: the dump is never uploaded."
canonical: https://intosql.com/mysql-to-postgresql
last_updated: 2026-10-15
---

# Convert MySQL to PostgreSQL

Drop a dump from mysqldump, mariadb-dump, mysqlpump, phpMyAdmin or Adminer, as .sql or .sql.gz, and download a PostgreSQL script with the same tables, keys, indexes and data. Every value PostgreSQL can't hold as it is, and every object it doesn't carry, is listed in the report.

## What changes from MySQL to PostgreSQL

Each row is a rule the converter applies, with the report line it writes when the rule changes something in your data.

| In MySQL | In PostgreSQL | Report line |
| --- | --- | --- |
| TINYINT(1), BOOL | boolean when every value is 0 or 1, else smallint | `bool_kept_numeric` |
| UNSIGNED integers | the next wider integer with a CHECK that it isn't negative; BIGINT UNSIGNED is bigint when every value fits, else numeric(20,0) | `unsigned_widened` |
| ZEROFILL and display widths | dropped | `zerofill_dropped` |
| FLOAT(M,D) | real or double precision, without the scale | `float_scale_dropped` |
| BIT(1), BIT(n) | boolean, bit(n) |  |
| zero, partial or impossible dates (0000-00-00, 2024-02-30) | NULL | `invalid_date_null` |
| NOT NULL columns with values that become NULL | NOT NULL dropped | `not_null_relaxed` |
| DEFAULT '0000-00-00' | no default | `zero_date_default` |
| TIME | time when every value is within a day, else interval | `time_as_interval` |
| YEAR | smallint with a CHECK of its range |  |
| CHAR(n) | varchar(n) |  |
| DATETIME | timestamp |  |
| TIMESTAMP | timestamptz in the time zone the dump states, UTC when it states none | `timestamp_zone_assumed` |
| ENUM | text with a CHECK of its values |  |
| ENUM holding '' (MySQL's error value) | the empty value allowed by the check | `enum_invalid_value` |
| ENUM or SET holding values outside its list | no CHECK | `enum_check_skipped` |
| SET | text[] with a CHECK that every element is in the list |  |
| JSON | jsonb; json when a value holds \u0000 | `json_not_jsonb` |
| NUL bytes in text | removed | `nul_removed` |
| text that isn't valid in its character set | the bad bytes as the replacement character | `invalid_utf8_replaced` |
| text decoded twice (ñ) | copied as it is | `mojibake_suspected` |
| AUTO_INCREMENT | GENERATED BY DEFAULT AS IDENTITY, with the counter set past the copied ids |  |
| MyISAM AUTO_INCREMENT on the second column of a key | one counter for the table | `myisam_group_autoincrement` |
| ON UPDATE CURRENT_TIMESTAMP | a BEFORE UPDATE trigger | `on_update_trigger` |
| generated columns | STORED | `virtual_as_stored` |
| generated columns with functions PostgreSQL lacks | a plain column | `generated_not_translated` |
| INVISIBLE columns | regular columns | `invisible_column` |
| CHECK constraints | translated, inline in CREATE TABLE |  |
| CHECK constraints with functions PostgreSQL lacks | not created | `check_skipped` |
| CHECK ... NOT ENFORCED | not created | `check_not_enforced` |
| defaults with functions PostgreSQL lacks | no default | `default_not_translated` |
| keys and indexes on a prefix, c(20) | an index on left(c, 20) | `prefix_unique` |
| FULLTEXT indexes | GIN on to_tsvector('simple', ...) | `fulltext_as_gin` |
| indexes on values over 2,000 bytes | hash | `index_on_wide_values` |
| SPATIAL indexes | not created without PostGIS | `spatial_index_skipped` |
| functional indexes PostgreSQL can't write | not created | `index_skipped` |
| foreign keys | after the data, NOT VALID, then validated with a warning on orphans |  |
| foreign keys to columns without a unique key | not created | `foreign_key_skipped` |
| table and column names | lowercase when that stays unique | `identifier_renamed` |
| names over 63 bytes | shortened with a hash | `name_shortened` |
| GEOMETRY, POINT, POLYGON... | WKB in bytea | `geometry_as_wkb` |
| INET4, INET6, UUID (MariaDB) | inet, uuid |  |
| BINARY, VARBINARY, BLOB | bytea |  |
| PARTITION BY | one table | `partitions_dropped` |
| WITH SYSTEM VERSIONING (MariaDB) | a table with the current rows | `system_versioning_dropped` |
| MyISAM, MEMORY and other engines | a regular table | `engine_dropped` |
| views, triggers, procedures, functions and events | listed in the report | `not_converted_objects` |
| dumps without CREATE TABLE | types read from the values | `no_schema` |
| geometries in a dump without CREATE TABLE | WKB in bytea | `geometry_inferred` |
| _ci collations | exact comparison | `case_insensitive_collation` |
| SET NAMES utf8 from utf8mb4 tables | converted, with a warning | `dump_lost_four_byte_chars` |
| dumps without SET NAMES whose text isn't UTF-8 | read as Latin-1 (Windows-1252) | `charset_guessed` |
| sjis, cp932, gbk, big5 dumps | read as UTF-8, with a warning | `unsupported_charset` |
| UNHEX('..'), CONVERT(b'..', UNSIGNED), CONVERT(".." USING x), ST_GeomFromText('..') in rows | evaluated |  |
| other expressions in rows | NULL | `expression_value` |
| DEFAULT in VALUES | the column's default value | `default_keyword_filled` |
| DEFAULT in VALUES without a literal default | NULL | `default_keyword_null` |
| REPLACE and INSERT IGNORE | plain rows | `replace_as_insert` |
| types IntoSQL doesn't know | text | `unknown_type` |
| values that don't fit their column | NULL | `invalid_value_null` |
| unreadable geometries | NULL | `bad_geometry` |
| CREATE SEQUENCE (MariaDB) | CREATE SEQUENCE with its current value |  |

## Load the script into PostgreSQL

The script creates the tables, copies the rows, then adds keys, indexes and foreign keys, so it loads fast and stops at the first error.

```shell
psql -v ON_ERROR_STOP=1 -d mydb -f mysql.postgresql.sql
```

Foreign keys are added as NOT VALID and then validated. If a row points to a row that doesn't exist, the script tells you which constraint, instead of failing half way.

For a large dump, create the database first and run the script with psql from the same machine as the server.

## Dump MySQL for the best result

The converter reads what the dump holds. These options keep more of your data in it:

- --default-character-set=utf8mb4 keeps emoji and other 4-byte characters, which a utf8 dump loses before the converter sees it.
- --hex-blob writes binary columns as hexadecimal, which survives copying between systems.
- --routines --triggers --events add those objects to the dump, so the report can list them.

## What isn't converted

Views, triggers, stored procedures, functions and events are written in each database's own language. They are listed in the report by name, never dropped silently, so you can rewrite them for PostgreSQL.

## Measured against pgloader

The same dumps, converted and loaded by IntoSQL and by pgloader 3.6.7 on one machine (a 10-core i9-10900K with a SATA SSD), in October 2026.

- A TINYINT(1) column holding 2, -1 and -5: pgloader loads true for each one, without a warning. IntoSQL keeps the numbers and the report says why.
- A dump with a default of 'it's', a table named with a backtick, an ENUM holding the empty value and a TIME over 24 hours: pgloader stops. IntoSQL converts it and loads it.
- What pgloader loads has lost its primary, unique and foreign keys and its identities. IntoSQL keeps every one.
- A 5.3 GB dump of 30 million rows: pgloader needs it restored into a running MySQL first, which takes 18 minutes, then runs out of memory with its defaults. IntoSQL converts the dump in 10 seconds, and psql loads the script in about 3 minutes.
- Across 476 dumps from MySQL 5.5 to 9 and MariaDB 10.3 to 12, loaded into PostgreSQL 13 and 18: no load errors, and no value changed without a line in the report.

## Questions about MySQL to PostgreSQL

### Is my dump uploaded?

No. The converter runs in your browser and reads the dump from your disk in pieces, so it never leaves your computer. The desktop app and the command line convert on your computer too.

### How large can the dump be?

The converter never holds the whole dump in memory: it reads the file twice, by ranges, and writes the script as it goes. Multi-gigabyte dumps convert in the browser, in the desktop app and with the command line.

### Does it read MariaDB dumps?

Yes. Dumps from mariadb-dump and from MariaDB 10.3 to 12 convert the same way, including MariaDB's INET6, INET4 and UUID columns.

### What happens to 0000-00-00 dates?

PostgreSQL has no zero date, so they become NULL and the column accepts NULL. The report says how many there were in each column.

### Are AUTO_INCREMENT values kept?

Yes. Each AUTO_INCREMENT column becomes an identity column, every row keeps its id, and the next id continues after the highest one, or after the table's AUTO_INCREMENT if that is higher.

### Does TINYINT(1) become boolean?

Only when every value in the column is 0 or 1. If it holds other numbers, such as 2 or -1, it stays a number and the report says why, so no value turns into true by accident.

### What about stored procedures and triggers?

They are listed in the report by name and not converted, because each database writes them in its own language. ON UPDATE CURRENT_TIMESTAMP is the exception: it becomes a PostgreSQL trigger.

### Which PostgreSQL versions does the script load on?

PostgreSQL 13 to 18. The same script loads on each of them, stopping at the first error if anything goes wrong.

## Other conversions to PostgreSQL

- [Convert Excel to PostgreSQL](https://intosql.com/excel-to-postgresql)
- [Convert CSV to PostgreSQL](https://intosql.com/csv-to-postgresql)
- [Convert JSON to PostgreSQL](https://intosql.com/json-to-postgresql)