Autogenerated migration misses changes to server_default and onupdate

[jankatins]

Describe the bug

If a SQLA model was already created and now the server_default or on_update changes/gets added, there is no migration created.

Expected behavior

I expect some alter column statement getting generated.

To Reproduce

before:

metadata_schema_name = "test_schema"
Base: Any = declarative_base(metadata=MetaData(schema=metadata_schema_name))
class Test(Base):
    __tablename__ = "_test"
    id = Column(BigInteger, primary_key=True, nullable=False)
    created = Column(DateTime(timezone=True), server_default=func.now(), nullable=False)
    last_updated = Column(DateTime(timezone=True), onupdate=func.now())

Alembic generates a migration for this:

    op.create_table('_test',
    sa.Column('id', sa.BigInteger(), nullable=False),
    sa.Column('created', sa.DateTime(timezone=True), server_default=sa.text('now()'), nullable=False),
    sa.Column('last_updated', sa.DateTime(timezone=True), nullable=True),
    sa.PrimaryKeyConstraint('id'),
    schema='test_schema'
    )

Changing the model to:

class Test(Base):
    __tablename__ = "_test"
    id = Column(BigInteger, primary_key=True, nullable=False)
    created = Column(DateTime(timezone=True), server_default=func.clock_timestamp(), nullable=False)
    last_updated = Column(DateTime(timezone=True), server_default=func.clock_timestamp(),
                          onupdate=func.clock_timestamp(), )

and running alembic revision --autogenerate -m "test2" will essentially create an empty migration (I have some autogenerated "CREATE SCHEMA IF NOT EXIST ... " in there). If I instead generate a complete new table (by simply changing the name so the old table gets removed and the new one gets created), I get the correct migration:

    op.create_table('_test2',
    sa.Column('id', sa.BigInteger(), nullable=False),
    sa.Column('created', sa.DateTime(timezone=True), server_default=sa.text('clock_timestamp()'), nullable=False),
    sa.Column('last_updated', sa.DateTime(timezone=True), server_default=sa.text('clock_timestamp()'), nullable=True),
    sa.PrimaryKeyConstraint('id'),
    schema='test_schema'
    )
    op.drop_table('_test', schema='test_schema')

(Update: I also just realized that no trigger is getting created for onupdate :-( Ouupsi, glad that I saw this...)

Error

None :-( Also no warnings relating to this (I've cycles in my tables and also some warnings re "expression-based index" due to a exclusion statement).

Also no warning that onupdate is not working on PG :-(

Versions.

• OS: Mac, latest version, intel
• Python: 3.11
• Alembic: alembic==1.9.4, alembic-utils==0.8.1
• SQLAlchemy: SQLAlchemy==2.0.4
• Database: PG 15
• DBAPI: psycopg2-binary==2.9.5

[CaselIT]

Hi,

That's working as expected. server default are not compared by default. See https://alembic.sqlalchemy.org/en/latest/api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.compare_server_default

Also no warning that onupdate is not working on PG :-(

not sure what you mean by this. onupdate is related to what sqlalchemy will do when issuing an update https://docs.sqlalchemy.org/en/20/core/metadata.html#sqlalchemy.schema.Column.params.onupdate

[adrianovieira]

Same issue here.

Installed packages:

• Alembic: 1.17.2
• SQLAlchemy: 2.0.44
• PostgreSQL psycopg[binary, pool]: 3.2.13

The update_at column is the focous of I think is an issue,

Excerpt of the migration script for Users model description:

• actual result (server_onupdate is missing)

sa.Column('updated_at', sa.DateTime(timezone=True), 
   server_default=sa.text('now()'), 
   nullable=False),

• expected result

sa.Column('updated_at', sa.DateTime(timezone=True), 
    server_default=sa.text('now()'), 
    server_onupdate=sa.text('now()'),
    nullable=False),

Would you happen to have any other tips to give?

Other infos:

Users model

from datetime import datetime
from sqlalchemy import DateTime, func
from sqlalchemy.orm import DeclarativeBase, Mapped, mapped_column

class Base(DeclarativeBase):
    pass

class Users(Base):
    __tablename__ = "users"

    id: Mapped[int] = mapped_column(primary_key=True)
    name: Mapped[str]

    created_at: Mapped[datetime] = mapped_column(
        DateTime(timezone=True),
        server_default=func.now(),
    )
    updated_at: Mapped[datetime] = mapped_column(
        DateTime(timezone=True),
        server_default=func.now(),

        # It doesn't matter if we use both parameters below or just one of them.
        # The results are the same:  missing the `server_onupdate` on the migration script.
        server_onupdate=func.now(),
        onupdate=func.now(),
    )

Full migration script

"""Users: create table

Revision ID: f98bb5629b6e
Revises:
Create Date: 2025-11-23 04:10:00.274823

"""

from typing import Sequence

import sqlalchemy as sa
from alembic import op

# revision identifiers, used by Alembic.
revision: str = "f98bb5629b6e"
down_revision: str | Sequence[str] | None = None
branch_labels: str | Sequence[str] | None = None
depends_on: str | Sequence[str] | None = None


def upgrade() -> None:
    """Upgrade schema."""
    # ### commands auto generated by Alembic - please adjust! ###
    op.create_table(
        "users",
        sa.Column("id", sa.Integer(), nullable=False),
        sa.Column("name", sa.String(), nullable=False),
        sa.Column(
            "created_at",
            sa.DateTime(timezone=True),
            server_default=sa.text("now()"),
            nullable=False,
        ),
        sa.Column(
            "updated_at",
            sa.DateTime(timezone=True),
            server_default=sa.text("now()"),
            nullable=False,
        ),
        sa.PrimaryKeyConstraint("id"),
    )
    # ### end Alembic commands ###


def downgrade() -> None:
    """Downgrade schema."""
    # ### commands auto generated by Alembic - please adjust! ###
    op.drop_table("users")
    # ### end Alembic commands ###

[CaselIT]

These is no ddl that's rendered by server_onupdate so that's not considered by alembic