In this page you will find information about the settings that are used on GitLab.com.
SSH host keys fingerprints
Below are the fingerprints for GitLab.com's SSH host keys.
GitLab.com sends emails from the
mg.gitlab.com domain via Mailgun and has
its own dedicated IP address (
Alternative SSH port
GitLab.com can be reached via a different SSH port for
~/.ssh/config is the following:
Host gitlab.com Hostname altssh.gitlab.com User git Port 443 PreferredAuthentications publickey IdentityFile ~/.ssh/gitlab
Below are the settings for GitLab Pages.
|Custom domains support||yes||no|
|TLS certificates support||yes||no|
The maximum size of your Pages site is regulated by the artifacts maximum size which is part of GitLab CI/CD.
Below are the current settings regarding GitLab CI/CD.
|Artifacts maximum size||1G||100M|
|Artifacts expiry time||kept forever||deleted after 30 days unless otherwise specified|
Repository size limit
The maximum size your Git repository is allowed to be, including LFS. If you are near or over the size limit, you can reduce your repository size with Git.
|Repository size including LFS||10G||Unlimited|
GitLab.com, CI/CD, and related services are deployed into Google Cloud Platform (GCP). Any IP based firewall can be configured by looking up all IP address ranges or CIDR blocks for GCP.
Static endpoints are being considered.
Shared Runners on GitLab.com run in autoscale mode and powered by Google Cloud Platform. Autoscaling means reduced waiting times to spin up CI/CD jobs, and isolated VMs for each project, thus maximizing security. They're free to use for public open source projects and limited to 2000 CI minutes per month per group for private projects. Read about all GitLab.com plans.
All your CI/CD jobs run on n1-standard-1 instances with 3.75GB of RAM, CoreOS and the latest Docker Engine installed. Instances provide 1 vCPU and 25GB of HDD disk space. The default region of the VMs is US East1. Each instance is used only for one job, this ensures any sensitive data left on the system can't be accessed by other people their CI jobs.
Jobs handled by the shared Runners on GitLab.com (
will be timed out after 3 hours, regardless of the timeout configured in a
project. Check the issues 4010 and 4070 for the reference.
Below are the shared Runners settings.
|GitLab Runner||Runner versions dashboard||-|
|Default Docker image||
The full contents of our
concurrent = X check_interval = 1 metrics_server = "X" sentry_dsn = "X" [[runners]] name = "docker-auto-scale" request_concurrency = X url = "https://gitlab.com/" token = "SHARED_RUNNER_TOKEN" executor = "docker+machine" environment = [ "DOCKER_DRIVER=overlay2" ] limit = X [runners.docker] image = "ruby:2.5" privileged = true [runners.machine] IdleCount = 20 IdleTime = 1800 OffPeakPeriods = ["* * * * * sat,sun *"] OffPeakTimezone = "UTC" OffPeakIdleCount = 5 OffPeakIdleTime = 1800 MaxBuilds = 1 MachineName = "srm-%s" MachineDriver = "digitalocean" MachineOptions = [ "digitalocean-image=X", "digitalocean-ssh-user=core", "digitalocean-region=nyc1", "digitalocean-size=s-2vcpu-2gb", "digitalocean-private-networking", "digitalocean-tags=shared_runners,gitlab_com", "engine-registry-mirror=http://INTERNAL_IP_OF_OUR_REGISTRY_MIRROR", "digitalocean-access-token=DIGITAL_OCEAN_ACCESS_TOKEN", ] [runners.cache] Type = "s3" BucketName = "runner" Insecure = true Shared = true ServerAddress = "INTERNAL_IP_OF_OUR_CACHE_SERVER" AccessKey = "ACCESS_KEY" SecretKey = "ACCESS_SECRET_KEY"
Google Cloud Platform
concurrent = X check_interval = 1 metrics_server = "X" sentry_dsn = "X" [[runners]] name = "docker-auto-scale" request_concurrency = X url = "https://gitlab.com/" token = "SHARED_RUNNER_TOKEN" executor = "docker+machine" environment = [ "DOCKER_DRIVER=overlay2" ] limit = X [runners.docker] image = "ruby:2.5" privileged = true [runners.machine] IdleCount = 20 IdleTime = 1800 OffPeakPeriods = ["* * * * * sat,sun *"] OffPeakTimezone = "UTC" OffPeakIdleCount = 5 OffPeakIdleTime = 1800 MaxBuilds = 1 MachineName = "srm-%s" MachineDriver = "google" MachineOptions = [ "google-project=PROJECT", "google-disk-size=25", "google-machine-type=n1-standard-1", "google-username=core", "google-tags=gitlab-com,srm", "google-use-internal-ip", "google-zone=us-east1-d", "google-machine-image=PROJECT/global/images/IMAGE", "engine-registry-mirror=http://INTERNAL_IP_OF_OUR_REGISTRY_MIRROR" ] [runners.cache] Type = "s3" BucketName = "runner" Insecure = true Shared = true ServerAddress = "INTERNAL_IP_OF_OUR_CACHE_SERVER" AccessKey = "ACCESS_KEY" SecretKey = "ACCESS_SECRET_KEY"
GitLab.com runs Sidekiq with arguments
and the following environment variables:
Periodically executed jobs by Sidekiq, to self-heal GitLab, do external synchronizations, run scheduled pipelines, etc.:
GitLab.com being a fairly large installation of GitLab means we have changed various PostgreSQL settings to better suit our needs. For example, we use streaming replication and servers in hot-standby mode to balance queries across different database servers.
The list of GitLab.com specific settings (and their defaults) is as follows:
|effective_cache_size||338688MB||Based on how much memory is available|
|shared_buffers||112896MB||Based on how much memory is available|
|shmall||30146560||Based on the server's capabilities|
|shmmax||123480309760||Based on the server's capabilities|
Some of these settings are in the process being adjusted. For example, the value
shared_buffers is quite high and as such we are looking into adjusting it.
More information on this particular change can be found at
https://gitlab.com/gitlab-com/infrastructure/issues/1555. An up to date list
of proposed changes can be found at
GitLab.com adjusts the memory limits for the unicorn-worker-killer gem.
GitLab.com at scale
In addition to the GitLab Enterprise Edition Omnibus install, GitLab.com uses the following applications and settings to achieve scale. All settings are located publicly available chef cookbooks.
We use Elasticsearch, logstash, and Kibana for part of our monitoring solution:
Prometheus complete our monitoring stack:
For the visualization of monitoring data:
Open source error tracking:
High Performance TCP/HTTP Load Balancer: