Putting a gateway in front of your providers adds a hop to your email path.
These are the questions worth settling before production mail goes through
it. Each answer is short, and it links to the page that has the detail.
Where Mailway doesn’t have an answer yet, the row says not yet.
| Ask |
The answer |
What happens to an SMTP message right after the 250? |
The 250 goes out once the message is written to the disk of Mailway’s SMTP relay and flushed there, so a crash or a power cut doesn’t lose it. Normally within seconds (longer during a burst) it becomes a database record and an archived copy. Until then it is on that one machine’s disk, which isn’t backed up: a disk lost in that window would lose what was on it, and only that. The path of a message |
| And over the HTTP API? |
The record and the request body are committed to the database before the 202 Accepted returns. The path of a message |
| What if Mailway can’t take the message? |
Over SMTP your mailer gets a 4xx reply, or no connection at all, and the message stays with your app. A mail server queues it and retries on its own; an app library such as nodemailer, PHPMailer, or Symfony Mailer reports an error and doesn’t retry. Over the HTTP API your client gets a non-2xx response with the error envelope, or a timeout. Check that your app really queues and retries: changing the SMTP host doesn’t add retries your app never had. SMTP reply codes · API errors |
| Can a recipient get a mail twice? |
Yes, in two documented cases. If a provider attempt’s answer is lost after the message went out, Mailway can’t tell whether the provider took it, and tries again. If a send has more recipients than the provider takes in one message (50 on Amazon SES), Mailway sends it in batches, and when one batch fails, the retry sends every batch again. Over the HTTP API, an Idempotency-Key makes your own retries safe: a repeated key returns the first send, even with a different body. Single point of failure · Idempotency |
| What if every provider refuses it? |
The mail becomes Exhausted. It shows in the console, a mail.failed webhook fires, and on Pro and up you can retry it, which starts the chain again. It never sits quietly in Pending. Failure semantics |
| Ask |
The answer |
| Is each provider ready on its own? |
Only if you test it. On each provider’s page in the console, Send test mail sends one mail straight through that account (admins only). To test the whole path through Mailway, pin a send to one provider with the HTTP API’s routing.provider field. Failover needs at least two providers on the project, on Pro and up. Per-send routing · Failover, precisely |
| Which providers report delivery events? |
Every supported provider except Gmail and Any SMTP server (which covers Microsoft 365, SparkPost, Elastic Email, and Zoho ZeptoMail), once you set up delivery tracking for that provider in the console. Without events, a mail’s last state is Accepted: the provider took it, and nothing confirms delivery. Supported providers |
| Is the archived copy what the recipient got? |
Not exactly. The archive holds what your app submitted. Most providers we tested change their copy before delivery: some change a header or two, some the HTML (tracking pixels, rewritten links). Mailway itself changes two things automatically. Above a provider’s recipient limit it replaces the visible recipient headers with To: undisclosed-recipients:;. It also adds provider instruction headers, which the provider reads and removes. What providers change · What happens to your message |
| How do I deduplicate webhook events? |
Webhooks are signed and delivered at least once; deduplicate on (type, data.mail.uid). Webhook delivery semantics |
| Will I notice a failing webhook endpoint? |
Each delivery is attempted six times over about 8.6 hours. An endpoint that keeps failing for 24 hours is turned off, and Mailway tells the team by mail and on the console’s Webhooks page. Events that happen while it is off are never sent to it, not even after you turn it back on, so fix it quickly. Webhook delivery semantics |
| Does a bounce on one provider protect the others? |
Yes, within a project: hard bounces and complaints suppress the address across every provider on it. Suppressions aren’t shared between projects. Suppressions |
| Ask |
The answer |
| How long is content kept, and what stays after? |
Content follows your plan’s window, or a shorter one a team admin sets for the team or a project. Delivery history, the envelope details (including subject and recipients), and the message fingerprint stay for the life of the account. Monthly bundles can keep content up to about a month longer. Data retention |
| Who on my team can see what? |
Four roles: admin, developer, finance, and member. Admins see every project. Developers and members see only the projects they are given, and finance sees billing only. Only admins can add or change provider accounts and their secrets. Concepts: team · Trust & security |
| How do I rotate or revoke a credential? |
To rotate, create a new one in the project’s Access section, move your app to it, then revoke the old one (admins and developers can do this). Each SMTP credential and API key belongs to one project, so a leaked one can only send through that project. API keys can also be created with an expiry date. Credentials & API keys |
| Where does mail live, and who processes it? |
Mailway’s servers are in Falkenstein, Germany, and archived content is in AWS S3 in Frankfurt. The trust page lists Mailway’s sub-processors with their roles. The providers you connect are your own vendors, so they aren’t on that list, and a mail goes wherever your provider takes it. Residency · Sub-processors |
| What does a certificate prove? |
That Mailway signed a record of the archived message bytes, the provider’s handoff, and the delivery events it lists, and that this record existed at the time on its timestamp. The timestamp dates the certificate, not the send. The check proves the certificate is intact and was signed by the key it carries. Mailway doesn’t publish its signing key yet, so the check alone doesn’t prove who issued it. Certificates are on Business and up. Proof-of-delivery certificates |
| Ask |
The answer |
| Can my traffic hit a limit? |
Free stops at its monthly limits and also has a daily limit. Paid plans keep sending past the monthly mail limit (soft overage, not billed during the private beta). Past the data-volume limit they get a grace margin, then new mail is refused (402 over the API, 550 over SMTP) until the month rolls over, you add an add-on, or you upgrade. Model your real message sizes, including attachments. Limits & quotas |
| How big can a message be? |
25 MB per message, over SMTP and the HTTP API. The HTTP API is stricter inside that: html and text up to 2 MB each, at most 20 attachments of up to 10 MB each and 25 MB together, and at most 50 recipients. Message limits |
| Is there more than one region? |
No. Mailway runs in one EU region (Falkenstein, Germany), with no multi-region failover. One region |
| Is there an uptime commitment? |
No. The terms provide the service “as is”, with no guaranteed uptime level or SLA. An Enterprise contract can include one. status.mailway.net shows whether Mailway is up right now; an independent monitoring service runs the checks. Terms · Service status |
| Ask |
The answer |
| How do I roll back? |
Over SMTP, point your mailer back at your provider’s SMTP server (host, port, credentials). Over the HTTP API, your app has to call your provider’s own API again, because Mailway’s send endpoint has its own request format. Your provider accounts never moved, so there is no DNS or reputation to migrate. If you leave |
| Can I take the record with me? |
Yes. A team admin can export everything under Team settings: every mail Mailway still stores, as Mailway stored it, and the delivery history as CSV files. The zip files can be downloaded for 7 days. Each mail’s raw source also downloads from the console while its content is kept. If you leave |
