summaryrefslogtreecommitdiff
path: root/doc/install/requirements.md
diff options
context:
space:
mode:
Diffstat (limited to 'doc/install/requirements.md')
-rw-r--r--doc/install/requirements.md97
1 files changed, 60 insertions, 37 deletions
diff --git a/doc/install/requirements.md b/doc/install/requirements.md
index 35586091f74..175dfc62096 100644
--- a/doc/install/requirements.md
+++ b/doc/install/requirements.md
@@ -69,7 +69,7 @@ so keep in mind that you need at least 4GB available before running GitLab. With
less memory GitLab will give strange errors during the reconfigure run and 500
errors during usage.
-- 1GB RAM + 3GB of swap is the absolute minimum but we strongly **advise against** this amount of memory. See the unicorn worker section below for more advice.
+- 1GB RAM + 3GB of swap is the absolute minimum but we strongly **advise against** this amount of memory. See the [unicorn worker section below](#unicorn-workers) for more advice.
- 2GB RAM + 2GB swap supports up to 100 users but it will be very slow
- **4GB RAM** is the **recommended** memory size for all installations and supports up to 100 users
- 8GB RAM supports up to 1,000 users
@@ -86,51 +86,42 @@ if your available memory changes.
Notice: The 25 workers of Sidekiq will show up as separate processes in your process overview (such as top or htop) but they share the same RAM allocation since Sidekiq is a multithreaded application. Please see the section below about Unicorn workers for information about many you need of those.
-## GitLab Runner
-
-We strongly advise against installing GitLab Runner on the same machine you plan
-to install GitLab on. Depending on how you decide to configure GitLab Runner and
-what tools you use to exercise your application in the CI environment, GitLab
-Runner can consume significant amount of available memory.
-
-Memory consumption calculations, that are available above, will not be valid if
-you decide to run GitLab Runner and the GitLab Rails application on the same
-machine.
-
-It is also not safe to install everything on a single machine, because of the
-[security reasons] - especially when you plan to use shell executor with GitLab
-Runner.
-
-We recommend using a separate machine for each GitLab Runner, if you plan to
-use the CI features.
-
-[security reasons]: https://gitlab.com/gitlab-org/gitlab-ci-multi-runner/blob/master/docs/security/index.md
-
-## Unicorn Workers
-
-It's possible to increase the amount of unicorn workers and this will usually help to reduce the response time of the applications and increase the ability to handle parallel requests.
+## Database
-For most instances we recommend using: CPU cores + 1 = unicorn workers.
-So for a machine with 2 cores, 3 unicorn workers is ideal.
+The server running the database should have _at least_ 5-10 GB of storage
+available, though the exact requirements depend on the size of the GitLab
+installation (e.g. the number of users, projects, etc).
-For all machines that have 2GB and up we recommend a minimum of three unicorn workers.
-If you have a 1GB machine we recommend to configure only two Unicorn workers to prevent excessive swapping.
+We currently support the following databases:
-To change the Unicorn workers when you have the Omnibus package please see [the Unicorn settings in the Omnibus GitLab documentation](https://gitlab.com/gitlab-org/omnibus-gitlab/blob/master/doc/settings/unicorn.md#unicorn-settings).
+- PostgreSQL (highly recommended)
+- MySQL/MariaDB (strongly discouraged, not all GitLab features are supported, no support for [MySQL/MariaDB GTID](https://mariadb.com/kb/en/mariadb/gtid/))
-## Database
+We highly recommend the use of PostgreSQL instead of MySQL/MariaDB as not all
+features of GitLab work with MySQL/MariaDB:
-We currently support the following databases:
+1. MySQL support for subgroups was [dropped with GitLab 9.3][post].
+ See [issue #30472][30472] for more information.
+1. GitLab Geo does [not support MySQL](https://docs.gitlab.com/ee/gitlab-geo/database.html#mysql-replication).
+1. [Zero downtime migrations][zero] do not work with MySQL
+1. GitLab [optimizes the loading of dashboard events](https://gitlab.com/gitlab-org/gitlab-ce/issues/31806) using [PostgreSQL LATERAL JOINs](https://blog.heapanalytics.com/postgresqls-powerful-new-join-type-lateral/).
+1. In general, SQL optimized for PostgreSQL may run much slower in MySQL due to
+ differences in query planners. For example, subqueries that work well in PostgreSQL
+ may not be [performant in MySQL](https://dev.mysql.com/doc/refman/5.7/en/optimizing-subqueries.html)
+1. We expect this list to grow over time.
-- PostgreSQL (recommended)
-- MySQL/MariaDB
+Existing users using GitLab with MySQL/MariaDB are advised to
+[migrate to PostgreSQL](../update/mysql_to_postgresql.md) instead.
-If you want to run the database separately, expect a size of about 1 MB per user.
+[30472]: https://gitlab.com/gitlab-org/gitlab-ce/issues/30472
+[zero]: ../update/README.md#upgrading-without-downtime
+[post]: https://about.gitlab.com/2017/06/22/gitlab-9-3-released/#dropping-support-for-subgroups-in-mysql
### PostgreSQL Requirements
-As of GitLab 9.0, PostgreSQL 9.6 is recommended. Lower versions of PostgreSQL
-may work but primary testing and developement takes place using PostgreSQL 9.6.
+As of GitLab 9.3, PostgreSQL 9.2 or newer is required, and earlier versions are
+not supported. We highly recommend users to use at least PostgreSQL 9.6 as this
+is the PostgreSQL version used for development and testing.
Users using PostgreSQL must ensure the `pg_trgm` extension is loaded into every
GitLab database. This extension can be enabled (using a PostgreSQL super user)
@@ -143,6 +134,18 @@ CREATE EXTENSION pg_trgm;
On some systems you may need to install an additional package (e.g.
`postgresql-contrib`) for this extension to become available.
+## Unicorn Workers
+
+It's possible to increase the amount of unicorn workers and this will usually help to reduce the response time of the applications and increase the ability to handle parallel requests.
+
+For most instances we recommend using: CPU cores + 1 = unicorn workers.
+So for a machine with 2 cores, 3 unicorn workers is ideal.
+
+For all machines that have 2GB and up we recommend a minimum of three unicorn workers.
+If you have a 1GB machine we recommend to configure only two Unicorn workers to prevent excessive swapping.
+
+To change the Unicorn workers when you have the Omnibus package please see [the Unicorn settings in the Omnibus GitLab documentation](https://gitlab.com/gitlab-org/omnibus-gitlab/blob/master/doc/settings/unicorn.md#unicorn-settings).
+
## Redis and Sidekiq
Redis stores all user sessions and the background task queue.
@@ -161,8 +164,28 @@ default settings.
If you would like to disable Prometheus and it's exporters or read more information
about it, check the [Prometheus documentation](../administration/monitoring/prometheus/index.md).
+## GitLab Runner
+
+We strongly advise against installing GitLab Runner on the same machine you plan
+to install GitLab on. Depending on how you decide to configure GitLab Runner and
+what tools you use to exercise your application in the CI environment, GitLab
+Runner can consume significant amount of available memory.
+
+Memory consumption calculations, that are available above, will not be valid if
+you decide to run GitLab Runner and the GitLab Rails application on the same
+machine.
+
+It is also not safe to install everything on a single machine, because of the
+[security reasons] - especially when you plan to use shell executor with GitLab
+Runner.
+
+We recommend using a separate machine for each GitLab Runner, if you plan to
+use the CI features.
+
+[security reasons]: https://gitlab.com/gitlab-org/gitlab-ci-multi-runner/blob/master/docs/security/index.md
+
## Supported web browsers
We support the current and the previous major release of Firefox, Chrome/Chromium, Safari and Microsoft browsers (Microsoft Edge and Internet Explorer 11).
-Each time a new browser version is released, we begin supporting that version and stop supporting the third most recent version. \ No newline at end of file
+Each time a new browser version is released, we begin supporting that version and stop supporting the third most recent version.