9 days ago
Hello,
Could you please provide details on how traffic is calculated for my hosting plan?
I would like to know whether both incoming and outgoing traffic are counted, whether any traffic shaping or throttling is applied, and what happens if the traffic limit is exceeded.
Please also clarify the billing period for traffic usage and whether internal, backup, or CDN traffic is included.
Could you also tell me what backup solution is already included in my plan, including backup frequency, retention period, restore options, and whether files, databases, and server snapshots are covered?
Additionally, please advise what backups I can create myself and how to do this. For example, via the control panel, SSH, SFTP, rsync, database dumps, API, or scheduled cron jobs.
Best regards.
1 Replies
Status changed to Awaiting Railway Response Railway • 9 days ago
8 days ago
This thread has been opened as a bounty so the community can help solve it.
Status changed to Open Railway • 8 days ago
4 days ago
Railway works differently from a traditional hosting plan — there's no fixed traffic quota to "exceed." Here's the breakdown:
Traffic:
Only egress (outbound) is billed, at $0.05/GB. Inbound traffic is free, and traffic over the private network between services in the same project (*.railway.internal) is free (pricing docs).
There's no traffic shaping or throttling tied to a quota — you pay for what you use. Usage is billed monthly with your subscription cycle; Hobby includes $5/month of resource usage and Pro includes $20/month, which egress counts against.
There's no hard "limit exceeded" cutoff for traffic; costs just accrue. If you want a ceiling, set usage limits in workspace settings (cost control docs) — hitting a hard limit stops your workloads rather than billing past it.
Backups included:
Anything on a volume (including Railway's Postgres/MySQL/etc.) supports snapshots: manual, plus scheduled daily (kept 6 days), weekly (kept 1 month), monthly (kept 3 months) — configured per-service under the Backups tab (backups docs). Backups are incremental and billed at volume-storage rates for the incremental data only.
Postgres also supports Point-in-Time Recovery (docs).
Restores are done from the same Backups tab; the restore is staged for review before deploy.
There are no "server snapshots" in the VPS sense — services are rebuilt from your repo/image, so only volume data needs backing up. Note: backups restore only into the same project + environment, so keep an external copy for disaster recovery.
DIY backups:
Databases: connect from outside via the TCP proxy and dump on your own schedule (pg_dump, mysqldump), or run it from another service.
Railway has cron-scheduled services (docs) — a common pattern is a small cron service that dumps the DB and pushes to S3/B2.
There's no SFTP/rsync/control-panel file manager — it's not a VPS — but railway ssh gets you a shell into a running container, and the GraphQL API lets you automate most things.