matrix-docker-ansible-deploy/docs/maintenance-synapse.md
Slavi Pantaleev d0cda27c97 Fix Synapse cache auto-tuning variables to use bytes, not KB
Fixup for https://github.com/spantaleev/matrix-docker-ansible-deploy/pull/3017

This reverts 1cd82cf068 and also multiplies results by `1024`
so as to pass bytes to Synapse, not KB (as done before).

1cd82cf068 was correctly documenting what we were doing (passing KB values),
but that's incorrect.

Synapse's Config Conventions
(https://matrix-org.github.io/synapse/latest/usage/configuration/config_documentation.html#config-conventions)
are supposed to clear it up, but they don't currently state what happens when you pass a plain number (without a unit suffix).

Thankfully, the source code tells us:
bc1db16086/synapse/config/_base.py (L181-L206)

> If an integer is provided it is treated as bytes and is unchanged.
>
> String byte sizes can have a suffix of ...
> No suffix is understood as a plain byte count.

We were previously passing strings, but that has been improved in 3d73ec887a.

Regardless, non-suffixed values seem to be treated as bytes by Synapse,
so this patch changes the variables to use bytes.

Moreover, we're moving from `matrix_synapse_memtotal_kb` to
`matrix_synapse_cache_size_calculations_memtotal_bytes` as working with
the base unit everywhere is preferrable.

Here, we also introduce 2 new variables to allow for the caps to be
tweaked:

- `matrix_synapse_cache_size_calculations_max_cache_memory_usage_cap_bytes`
- `matrix_synapse_cache_size_calculations_target_cache_memory_usage_cap_bytes`
2024-02-14 13:39:40 +02:00

7.4 KiB

Synapse maintenance

This document shows you how to perform various maintenance tasks related to the Synapse chat server.

Table of contents:

Purging old data with the Purge History API

You can use the Purge History API to delete old messages on a per-room basis. This is destructive (especially for non-federated rooms), because it means people will no longer have access to history past a certain point.

To make use of this Synapse Admin API, you'll need an admin access token first. Refer to the documentation on how to obtain an access token.

Synapse's Admin API is not exposed to the internet by default, following official Synapse reverse-proxying recommendations. To expose it you will need to add matrix_synapse_container_labels_public_client_synapse_admin_api_enabled: true to your vars.yml file.

Follow the Purge History API documentation page for the actual purging instructions.

After deleting data, you may wish to run a FULL Postgres VACUUM.

Compressing state with rust-synapse-compress-state

rust-synapse-compress-state can be used to optimize some _state tables used by Synapse. If your server participates in large rooms this is the most effective way to reduce the size of your database.

Note: besides running the rust-synapse-compress-state tool manually, you can also enable its synapse-auto-compressor tool by Setting up synapse-auto-compressor. The automatic tool will run on a schedule every day and you won't have to compress state manually ever again.

rust-synapse-compress-state should be safe to use (even when Synapse is running), but it's always a good idea to make Postgres backups first.

To ask the playbook to run rust-synapse-compress-state, execute:

ansible-playbook -i inventory/hosts setup.yml --tags=rust-synapse-compress-state

By default, all rooms with more than 100000 state group rows will be compressed. If you need to adjust this, pass: --extra-vars='matrix_synapse_rust_synapse_compress_state_min_state_groups_required=SOME_NUMBER_HERE' to the command above.

After state compression, you may wish to run a FULL Postgres VACUUM.

Browse and manipulate the database

When the Synapse Admin API and the other tools do not provide a more convenient way, having a look at synapse's postgresql database can satisfy a lot of admins' needs.

Editing the database manually is not recommended or supported by the Synapse developers. If you are going to do so you should make a database backup.

First, set up an SSH tunnel to your matrix server (skip if it is your local machine):

# you may replace 1799 with an arbitrary port unbound on both machines
ssh -L 1799:localhost:1799 matrix.DOMAIN

Then start up an ephemeral adminer container on the Matrix server, connecting it to the matrix network and linking the postgresql container:

docker run --rm --publish 1799:8080 --link matrix-postgres --net matrix adminer

You should then be able to browse the adminer database administration GUI at http://localhost:1799/ after entering your DB credentials (found in the host_vars or on the server in {{matrix_synapse_config_dir_path}}/homeserver.yaml under database.args)

⚠️ Be very careful with this, there is no undo for impromptu DB operations.

Make Synapse faster

Synapse's presence feature which tracks which users are online and which are offline can use a lot of processing power. You can disable presence by adding matrix_synapse_presence_enabled: false to your vars.yml file.

If you have enough compute resources (CPU & RAM), you can make Synapse better use of them by enabling load-balancing with workers.

Tuning your PostgreSQL database could also improve Synapse performance. The playbook tunes the integrated Postgres database automatically, but based on your needs you may wish to adjust tuning variables manually. If you're using an external Postgres database, you will aslo need to tune Postgres manually.

Tuning caches and cache autotuning

Tuning Synapse's cache factor is useful for performance increases but also as part of controlling Synapse's memory use. Use the variable matrix_synapse_caches_global_factor to set the cache factor as part of this process.

The playbook defaults the global cache factor to a large value (e.g. 10). A smaller value (e.g. 0.5) will decrease the amount used for caches, but will not necessarily decrease RAM usage as a whole.

Tuning the cache factor is useful only to a limited degree (as its crude to do in isolation) and therefore users who are tuning their cache factor should likely look into tuning autotune variables as well (see below).

Cache autotuning is enabled by default and controlled via the following variables:

  • matrix_synapse_cache_autotuning_max_cache_memory_usage - defaults to 1/8 of total RAM with a cap of 2GB; values are specified in bytes
  • matrix_synapse_cache_autotuning_target_cache_memory_usage - defaults to 1/16 of total RAM with a cap of 1GB; values are specified in bytes
  • matrix_synapse_cache_autotuning_min_cache_ttl - defaults to 30s

You can learn more about cache-autotuning and the global cache factor settings in the Synapse's documentation on caches and associated values.

To disable cache auto-tuning, unset all values:

matrix_synapse_cache_autotuning_max_cache_memory_usage: ''
matrix_synapse_cache_autotuning_target_cache_memory_usage: ''
matrix_synapse_cache_autotuning_min_cache_ttl: ''

Users who wish to lower Synapse's RAM footprint should look into lowering the global cache factor and tweaking the autotune variables (or disabling auto-tuning). If your cache factor is too low for a given auto tune setting your caches will not reach autotune thresholds and autotune won't be able to do its job. Therefore, when auto-tuning is enabled (which it is by default), it's recommended to have your cache factor be large.

See also How do I optimize this setup for a low-power server?.