Volumes

Volumes are a feature that enables persistent data for services on Railway.

How it works

When mounting a volume to a service, a volume is made available to the service on the specified mount path.

Size Limits

Volumes have a default size based on the subscription plan.

• Free and Trial plans: 0.5GB
• Hobby plans: 5GB
• Pro and team plans: 50GB

Volumes can be "Grown" after upgrading to a different plan.

Pro users and above can self-serve to increase their volume up to 250 GB.

For Pro and above users, please reach out to us on our Central Station if you need more than 250GB. Enterprise users with $2,000/month committed spend can also use Slack.

Volume Limits per Project

Each project has a maximum number of volumes that can be created, based on your subscription plan:

Plan	Maximum Volumes per Project
Free	1
Trial	3
Hobby	10
Pro	20

Volumes created for the same service in another environment do not count towards this limit.

Pro users with substantial usage requirements may be eligible for increased volume limits. Please reach out to us on our Central Station to discuss your project's needs.

I/O Specifications

Volumes have the following I/O characteristics:

• Read IOPS: 3,000 operations per second
• Write IOPS: 3,000 operations per second

These specifications apply to all volume sizes and subscription plans.

Pricing

Volumes are billed at a rate per GB / minutely, and invoiced monthly. You can find specific per-minute pricing here.

You are only charged for the amount of storage used by your volumes. Each volume requires approx 2-3% of the total storage to store metadata about the filesystem, so a new volume will always start with some used amount of space used depending on the size.

Backups

Services with volumes support manual and automated backups, backups are covered in the backups reference guide.

Caveats

Here are some limitations of which we are currently aware:

• Each service can only have a single volume
• Replicas cannot be used with volumes
• There is no built-in S/FTP support
• To prevent data corruption, we prevent multiple deployments from being active and mounted to the same service. This means that there will be a small amount of downtime when re-deploying a service that has a volume attached, even if there is a healthcheck endpoint configured
• Down-sizing a volume is not currently supported, but increasing size is supported
• When resizing a volume, all deployments must be taken offline to prevent data corruption
• There is no file browser, or direct file download. To access your files, you must do so via the attached service's mount point
• Docker images that run as a non-root UID by default will have permissions issues when performing operations within an attached volume. If you are affected by this, you can set RAILWAY_RUN_UID=0 environment variable in your service.

Support

Refer to the guide on how to use volumes for more details on how to use the feature.