ankane/pgslice

README

pgslice

Postgres partitioning as easy as pie. Works great for both new and existing tables, with zero downtime and minimal app changes. No need to install anything on your database server. Archive older data on a rolling basis to keep your database size under control.

🍊 Battle-tested at Instacart

Install

pgslice is a command line tool. To install, run:

gem install pgslice

This will give you the pgslice command. If installation fails, you may need to install dependencies.

You can also install it with Homebrew or Docker.

Steps

Ensure the table you want to partition has been created. We’ll refer to this as <table>.
Specify your database credentials

export PGSLICE_URL=postgres://localhost/myapp_development

Create an intermediate table

pgslice prep <table> <column> <period>

The column should be a timestamp, timestamptz, or date column and period can be day, month, or year.

This creates a partitioned table named <table>_intermediate using range partitioning.

Add partitions to the intermediate table

pgslice add_partitions <table> --intermediate --past 3 --future 3

Use the --past and --future options to control the number of partitions.

Optional, for tables with data - Fill the partitions in batches with data from the original table

pgslice fill <table>

Use the --batch-size and --sleep options to control the speed (defaults to 10000 and 0 respectively)

To sync data across different databases, check out pgsync.

Analyze tables

pgslice analyze <table>

Swap the intermediate table with the original table

pgslice swap <table>

The original table is renamed <table>_retired and the intermediate table is renamed <table>.

Fill the rest (rows inserted between the first fill and the swap)

pgslice fill <table> --swapped

Back up the retired table with a tool like pg_dump and drop it

pg_dump -c -Fc -t <table>_retired $PGSLICE_URL > <table>_retired.dump
psql -c "DROP TABLE <table>_retired" $PGSLICE_URL

Sample Output

pgslice prints the SQL commands that were executed on the server. To print without executing, use the --dry-run option.

pgslice prep visits created_at month

BEGIN;

CREATE TABLE "public"."visits_intermediate" (LIKE "public"."visits" INCLUDING DEFAULTS INCLUDING CONSTRAINTS INCLUDING STORAGE INCLUDING COMMENTS INCLUDING STATISTICS INCLUDING GENERATED INCLUDING COMPRESSION) PARTITION BY RANGE ("created_at");

CREATE INDEX ON "public"."visits_intermediate" USING btree ("created_at");

COMMENT ON TABLE "public"."visits_intermediate" is 'column:created_at,period:month,cast:date,version:3';

COMMIT;

pgslice add_partitions visits --intermediate --past 1 --future 1

BEGIN;

CREATE TABLE "public"."visits_202601" PARTITION OF "public"."visits_intermediate" FOR VALUES FROM ('2026-01-01') TO ('2026-02-01');

ALTER TABLE "public"."visits_202601" ADD PRIMARY KEY ("id");

CREATE TABLE "public"."visits_202602" PARTITION OF "public"."visits_intermediate" FOR VALUES FROM ('2026-02-01') TO ('2026-03-01');

ALTER TABLE "public"."visits_202602" ADD PRIMARY KEY ("id");

CREATE TABLE "public"."visits_202603" PARTITION OF "public"."visits_intermediate" FOR VALUES FROM ('2026-03-01') TO ('2026-04-01');

ALTER TABLE "public"."visits_202603" ADD PRIMARY KEY ("id");

COMMIT;

pgslice fill visits

/* 1 of 3 */
INSERT INTO "public"."visits_intermediate" ("id", "user_id", "ip", "created_at")
    SELECT "id", "user_id", "ip", "created_at" FROM "public"."visits"
    WHERE "id" > 0 AND "id" <= 10000 AND "created_at" >= '2026-01-01'::date AND "created_at" < '2026-04-01'::date

/* 2 of 3 */
INSERT INTO "public"."visits_intermediate" ("id", "user_id", "ip", "created_at")
    SELECT "id", "user_id", "ip", "created_at" FROM "public"."visits"
    WHERE "id" > 10000 AND "id" <= 20000 AND "created_at" >= '2026-01-01'::date AND "created_at" < '2026-04-01'::date

/* 3 of 3 */
INSERT INTO "public"."visits_intermediate" ("id", "user_id", "ip", "created_at")
    SELECT "id", "user_id", "ip", "created_at" FROM "public"."visits"
    WHERE "id" > 20000 AND "id" <= 30000 AND "created_at" >= '2026-01-01'::date AND "created_at" < '2026-04-01'::date

pgslice analyze visits

ANALYZE VERBOSE "public"."visits_202601";

ANALYZE VERBOSE "public"."visits_202602";

ANALYZE VERBOSE "public"."visits_202603";

ANALYZE VERBOSE "public"."visits_intermediate";

pgslice swap visits

BEGIN;

SET LOCAL lock_timeout = '5s';

ALTER TABLE "public"."visits" RENAME TO "visits_retired";

ALTER TABLE "public"."visits_intermediate" RENAME TO "visits";

ALTER SEQUENCE "public"."visits_id_seq" OWNED BY "public"."visits"."id";

COMMIT;

Adding Partitions

To add partitions, use:

pgslice add_partitions <table> --future 3

Add this as a cron job to create a new partition each day, month, or year.

# day
0 0 * * * pgslice add_partitions <table> --future 3 --url ...

# month
0 0 1 * * pgslice add_partitions <table> --future 3 --url ...

# year
0 0 1 1 * pgslice add_partitions <table> --future 3 --url ...

Add a monitor to ensure partitions are being created.

SELECT 1 FROM
    pg_catalog.pg_class c
INNER JOIN
    pg_catalog.pg_namespace n ON n.oid = c.relnamespace
WHERE
    c.relkind = 'r' AND
    n.nspname = 'public' AND
    c.relname = '<table>_' || to_char(NOW() + INTERVAL '3 days', 'YYYYMMDD')
    -- for months, use to_char(NOW() + INTERVAL '3 months', 'YYYYMM')
    -- for years, use to_char(NOW() + INTERVAL '3 years', 'YYYY')

Archiving Partitions

Back up and drop older partitions each day, month, or year.

pg_dump -c -Fc -t <table>_202601 $PGSLICE_URL > <table>_202601.dump
psql -c "DROP TABLE <table>_202601" $PGSLICE_URL

If you use Amazon S3 for backups, s3cmd is a nice tool.

s3cmd put <table>_202601.dump s3://<s3-bucket>/<table>_202601.dump

Schema Updates

Once a table is partitioned, make schema updates on the master table only (not partitions). This includes adding, removing, and modifying columns, as well as adding and removing indexes and foreign keys.

Additional Commands

To undo prep (which will delete partitions), use:

pgslice unprep <table>

To undo swap, use:

pgslice unswap <table>

Additional Options

Set the tablespace when adding partitions

pgslice add_partitions <table> --tablespace fastspace

App Considerations

This set up allows you to read and write with the original table name with no knowledge it’s partitioned. However, there are a few things to be aware of.

Reads

When possible, queries should include the column you partition on to limit the number of partitions the database needs to check. For instance, if you partition on created_at, try to include it in queries:

SELECT * FROM
    visits
WHERE
    user_id = 123 AND
    -- for performance
    created_at >= '2026-01-01' AND created_at < '2026-01-02'

For this to be effective, ensure constraint_exclusion is set to partition (the default value) or on.

SHOW constraint_exclusion;

Frameworks

Rails

Specify the primary key for partitioned models to ensure it’s returned.

class Visit < ApplicationRecord
  self.primary_key = "id"
end

Other Frameworks

Please submit a PR if additional configuration is needed.

One Off Tasks

You can also use pgslice to reduce the size of a table without partitioning by creating a new table, filling it with a subset of records, and swapping it in.

pgslice prep <table> --no-partition
pgslice fill <table> --where "id > 1000" # use any conditions
pgslice analyze <table>
pgslice swap <table>
pgslice fill <table> --where "id > 1000" --swapped

Triggers

Triggers aren’t copied from the original table. You can set up triggers on the intermediate table if needed.

Data Protection

Always make sure your connection is secure when connecting to a database over a network you don’t fully trust. Your best option is to connect over SSH or a VPN. Another option is to use sslmode=verify-full. If you don’t do this, your database credentials can be compromised.

Additional Installation Methods

Homebrew

With Homebrew, you can use:

brew install pgslice

Docker

Get the Docker image with:

docker pull ankane/pgslice
alias pgslice="docker run --rm -e PGSLICE_URL ankane/pgslice"

This will give you the pgslice command.

Dependencies

If installation fails, your system may be missing Ruby or libpq.

On Mac, run:

brew install libpq

On Ubuntu, run:

sudo apt-get install ruby-dev libpq-dev build-essential

Upgrading

Run:

gem install pgslice

To use master, run:

gem install specific_install
gem specific_install https://github.com/ankane/pgslice.git

Reference

PostgreSQL Manual

Related Projects

Also check out:

Dexter - The automatic indexer for Postgres
PgHero - A performance dashboard for Postgres
pgsync - Sync Postgres data to your local machine

History

View the changelog

Contributing

Everyone is encouraged to help improve this project. Here are a few ways you can help:

Report bugs
Fix bugs and submit pull requests
Write, clarify, or fix documentation
Suggest or add new features

To get started with development:

git clone https://github.com/ankane/pgslice.git
cd pgslice
bundle install
createdb pgslice_test
bundle exec rake test

To test against different versions of Postgres with Docker, use:

docker run -p=8000:5432 postgres:16
TZ=Etc/UTC PGSLICE_URL=postgres://postgres@localhost:8000/postgres bundle exec rake

On Mac, you must use Docker Desktop for the port mapping to localhost to work.

README

pgslice

🍊 Battle-tested at Instacart

Install

pgslice is a command line tool. To install, run:

gem install pgslice

This will give you the pgslice command. If installation fails, you may need to install dependencies.

You can also install it with Homebrew or Docker.

Steps

Ensure the table you want to partition has been created. We’ll refer to this as <table>.
Specify your database credentials

export PGSLICE_URL=postgres://localhost/myapp_development

Create an intermediate table

pgslice prep <table> <column> <period>

The column should be a timestamp, timestamptz, or date column and period can be day, month, or year.

This creates a partitioned table named <table>_intermediate using range partitioning.

Add partitions to the intermediate table

pgslice add_partitions <table> --intermediate --past 3 --future 3

Use the --past and --future options to control the number of partitions.

Optional, for tables with data - Fill the partitions in batches with data from the original table

pgslice fill <table>

Use the --batch-size and --sleep options to control the speed (defaults to 10000 and 0 respectively)

To sync data across different databases, check out pgsync.

Analyze tables

pgslice analyze <table>

Swap the intermediate table with the original table

pgslice swap <table>

The original table is renamed <table>_retired and the intermediate table is renamed <table>.

Fill the rest (rows inserted between the first fill and the swap)

pgslice fill <table> --swapped

Back up the retired table with a tool like pg_dump and drop it

pg_dump -c -Fc -t <table>_retired $PGSLICE_URL > <table>_retired.dump
psql -c "DROP TABLE <table>_retired" $PGSLICE_URL

Sample Output

pgslice prints the SQL commands that were executed on the server. To print without executing, use the --dry-run option.

pgslice prep visits created_at month

BEGIN;

CREATE TABLE "public"."visits_intermediate" (LIKE "public"."visits" INCLUDING DEFAULTS INCLUDING CONSTRAINTS INCLUDING STORAGE INCLUDING COMMENTS INCLUDING STATISTICS INCLUDING GENERATED INCLUDING COMPRESSION) PARTITION BY RANGE ("created_at");

CREATE INDEX ON "public"."visits_intermediate" USING btree ("created_at");

COMMENT ON TABLE "public"."visits_intermediate" is 'column:created_at,period:month,cast:date,version:3';

COMMIT;

pgslice add_partitions visits --intermediate --past 1 --future 1

BEGIN;

CREATE TABLE "public"."visits_202601" PARTITION OF "public"."visits_intermediate" FOR VALUES FROM ('2026-01-01') TO ('2026-02-01');

ALTER TABLE "public"."visits_202601" ADD PRIMARY KEY ("id");

CREATE TABLE "public"."visits_202602" PARTITION OF "public"."visits_intermediate" FOR VALUES FROM ('2026-02-01') TO ('2026-03-01');

ALTER TABLE "public"."visits_202602" ADD PRIMARY KEY ("id");

CREATE TABLE "public"."visits_202603" PARTITION OF "public"."visits_intermediate" FOR VALUES FROM ('2026-03-01') TO ('2026-04-01');

ALTER TABLE "public"."visits_202603" ADD PRIMARY KEY ("id");

COMMIT;

pgslice fill visits

/* 1 of 3 */
INSERT INTO "public"."visits_intermediate" ("id", "user_id", "ip", "created_at")
    SELECT "id", "user_id", "ip", "created_at" FROM "public"."visits"
    WHERE "id" > 0 AND "id" <= 10000 AND "created_at" >= '2026-01-01'::date AND "created_at" < '2026-04-01'::date

/* 2 of 3 */
INSERT INTO "public"."visits_intermediate" ("id", "user_id", "ip", "created_at")
    SELECT "id", "user_id", "ip", "created_at" FROM "public"."visits"
    WHERE "id" > 10000 AND "id" <= 20000 AND "created_at" >= '2026-01-01'::date AND "created_at" < '2026-04-01'::date

/* 3 of 3 */
INSERT INTO "public"."visits_intermediate" ("id", "user_id", "ip", "created_at")
    SELECT "id", "user_id", "ip", "created_at" FROM "public"."visits"
    WHERE "id" > 20000 AND "id" <= 30000 AND "created_at" >= '2026-01-01'::date AND "created_at" < '2026-04-01'::date

pgslice analyze visits

ANALYZE VERBOSE "public"."visits_202601";

ANALYZE VERBOSE "public"."visits_202602";

ANALYZE VERBOSE "public"."visits_202603";

ANALYZE VERBOSE "public"."visits_intermediate";

pgslice swap visits

BEGIN;

SET LOCAL lock_timeout = '5s';

ALTER TABLE "public"."visits" RENAME TO "visits_retired";

ALTER TABLE "public"."visits_intermediate" RENAME TO "visits";

ALTER SEQUENCE "public"."visits_id_seq" OWNED BY "public"."visits"."id";

COMMIT;

Adding Partitions

To add partitions, use:

pgslice add_partitions <table> --future 3

Add this as a cron job to create a new partition each day, month, or year.

# day
0 0 * * * pgslice add_partitions <table> --future 3 --url ...

# month
0 0 1 * * pgslice add_partitions <table> --future 3 --url ...

# year
0 0 1 1 * pgslice add_partitions <table> --future 3 --url ...

Add a monitor to ensure partitions are being created.

SELECT 1 FROM
    pg_catalog.pg_class c
INNER JOIN
    pg_catalog.pg_namespace n ON n.oid = c.relnamespace
WHERE
    c.relkind = 'r' AND
    n.nspname = 'public' AND
    c.relname = '<table>_' || to_char(NOW() + INTERVAL '3 days', 'YYYYMMDD')
    -- for months, use to_char(NOW() + INTERVAL '3 months', 'YYYYMM')
    -- for years, use to_char(NOW() + INTERVAL '3 years', 'YYYY')

Archiving Partitions

Back up and drop older partitions each day, month, or year.

pg_dump -c -Fc -t <table>_202601 $PGSLICE_URL > <table>_202601.dump
psql -c "DROP TABLE <table>_202601" $PGSLICE_URL

If you use Amazon S3 for backups, s3cmd is a nice tool.

s3cmd put <table>_202601.dump s3://<s3-bucket>/<table>_202601.dump

Schema Updates

Additional Commands

To undo prep (which will delete partitions), use:

pgslice unprep <table>

To undo swap, use:

pgslice unswap <table>

Additional Options

Set the tablespace when adding partitions

pgslice add_partitions <table> --tablespace fastspace

App Considerations

This set up allows you to read and write with the original table name with no knowledge it’s partitioned. However, there are a few things to be aware of.

Reads

SELECT * FROM
    visits
WHERE
    user_id = 123 AND
    -- for performance
    created_at >= '2026-01-01' AND created_at < '2026-01-02'

For this to be effective, ensure constraint_exclusion is set to partition (the default value) or on.

SHOW constraint_exclusion;

Frameworks

Rails

Specify the primary key for partitioned models to ensure it’s returned.

class Visit < ApplicationRecord
  self.primary_key = "id"
end

Other Frameworks

Please submit a PR if additional configuration is needed.

One Off Tasks

You can also use pgslice to reduce the size of a table without partitioning by creating a new table, filling it with a subset of records, and swapping it in.

pgslice prep <table> --no-partition
pgslice fill <table> --where "id > 1000" # use any conditions
pgslice analyze <table>
pgslice swap <table>
pgslice fill <table> --where "id > 1000" --swapped

Triggers

Triggers aren’t copied from the original table. You can set up triggers on the intermediate table if needed.

Data Protection

Additional Installation Methods

Homebrew

With Homebrew, you can use:

brew install pgslice

Docker

Get the Docker image with:

docker pull ankane/pgslice
alias pgslice="docker run --rm -e PGSLICE_URL ankane/pgslice"

This will give you the pgslice command.

Dependencies

If installation fails, your system may be missing Ruby or libpq.

On Mac, run:

brew install libpq

On Ubuntu, run:

sudo apt-get install ruby-dev libpq-dev build-essential

Upgrading

Run:

gem install pgslice

To use master, run:

gem install specific_install
gem specific_install https://github.com/ankane/pgslice.git

Reference

PostgreSQL Manual

Related Projects

Also check out:

Dexter - The automatic indexer for Postgres
PgHero - A performance dashboard for Postgres
pgsync - Sync Postgres data to your local machine

History

View the changelog

Contributing

Everyone is encouraged to help improve this project. Here are a few ways you can help:

Report bugs
Fix bugs and submit pull requests
Write, clarify, or fix documentation
Suggest or add new features

To get started with development:

git clone https://github.com/ankane/pgslice.git
cd pgslice
bundle install
createdb pgslice_test
bundle exec rake test

To test against different versions of Postgres with Docker, use:

docker run -p=8000:5432 postgres:16
TZ=Etc/UTC PGSLICE_URL=postgres://postgres@localhost:8000/postgres bundle exec rake

On Mac, you must use Docker Desktop for the port mapping to localhost to work.

ankane/pgslice(opens in a new tab)

README

pgslice

Install

Steps

Sample Output

Adding Partitions

Archiving Partitions

Schema Updates

Additional Commands

Additional Options

App Considerations

Reads

Frameworks

Rails

Other Frameworks

One Off Tasks

Triggers

Data Protection

Additional Installation Methods

Homebrew

Docker

Dependencies

Upgrading

Reference

Related Projects

History

Contributing

README

pgslice

Install

Steps

Sample Output

Adding Partitions

Archiving Partitions

Schema Updates

Additional Commands

Additional Options

App Considerations

Reads

Frameworks

Rails

Other Frameworks

One Off Tasks

Triggers

Data Protection

Additional Installation Methods

Homebrew

Docker

Dependencies

Upgrading

Reference

Related Projects

History

Contributing