Reliability and continuity

Reliability and continuity

RentalTide is the system of record for your bookings, payments, and operations. We design it so a single bad day — at AWS, at your dock, or at our office — does not turn into a bad week for your business. This page covers the broad strokes of what we do to keep the platform running and what your team can do when something does go wrong.

For deeper technical detail (regions, services, recovery time objectives), see the technical appendix.

What we protect against

Where your data lives

RentalTide runs entirely on Amazon Web Services. We do not store customer data on laptops, in our office, or on any device you cannot access from a browser. That means:

• A staff laptop being lost, stolen, or destroyed has no effect on your bookings, customers, or payments.
• A fire or flood at your location has no effect on your data. Sign in from a phone, a hotel laptop, or a tablet and you are back online.
• We do not need to ship hardware to recover from any disaster.

Your primary database runs in AWS's us-east-1 region in Northern Virginia with synchronous replication across three independent data centres (called Availability Zones). Continuous backups are taken automatically, and we keep point-in-time recovery for the last seven days plus daily snapshots for thirty-five days.

What happens if AWS has a bad day

AWS designs each region to tolerate the loss of one data centre with no customer impact. We rely on that. In the rare cases where an entire region has degraded service:

1. Detection — automated monitors alert RentalTide on-call within minutes.
2. Communication — we update the status page at status.rentaltide.com and post in your support channel.
3. Triage — for partial regional outages (one service degraded, others fine), we wait for AWS to recover. AWS regional incidents are typically resolved in under two hours.
4. Failover — for a full sustained regional outage (very rare, last meaningful one was December 2021), we restore from cross-region backups into us-west-2. Your data is preserved; the site comes back up within the recovery window in the technical appendix.

While we recover, your team can still do a lot — see the offline procedures below.

What happens if your internet drops

This is the most common outage you will actually experience. RentalTide is a web app, and if the dock has no internet, neither does the booking widget or the admin panel. Here is what to do.

Take walk-ups on paper

Keep a printed walk-up sheet on the counter with these fields:

• Customer first name, last name, phone, email
• Asset type and quantity
• Start time and intended return time
• Signature line for liability waiver
• Payment method (cash, card terminal direct, "bill me later")

Photos of the driver's licence and signed waiver work fine on a phone. Enter them in RentalTide once internet is restored.

Keep your card reader running

Your Stripe terminal can complete payments without RentalTide as long as it has cellular or hotspot connectivity. The transaction will still settle correctly. Reconcile it against the paper sheets when you re-sync.

Use a hotspot

A phone with cellular service is enough to get the admin panel back up on one device. Use the Today's Hub view on that device as your single source of truth until the dock connection returns.

Don't worry about the booking widget

Customers trying to book online during your local outage just see a normal "site unreachable" message from their browser. They will retry. We do not lose their attempts — our analytics show the booking widget itself remains available globally even when individual locations are offline.

What happens if our servers crash

Servers fail. Bad deploys ship. We plan for both:

• Automatic restart. If a server process crashes, our container orchestrator restarts it within seconds and shifts traffic to healthy instances.
• Health checks. Every server reports a heartbeat every few seconds. Any instance that fails its health check is removed from the rotation automatically.
• Bad deploy rollback. New code is deployed to a subset of servers first. If error rates spike, the deploy halts and the previous version takes back over. Total exposure on a bad deploy is usually under five minutes.
• Status page. Major incidents are posted at status.rentaltide.com with regular updates until resolution.

If you see errors that persist for more than a few minutes, reach out in your support channel. We will already be on it, but it helps us confirm scope.

What happens if there's a fire, flood, or worse

The honest answer: your data and bookings are unaffected. Everything that matters is in the cloud, replicated across three data centres. There is no scenario short of a regional AWS catastrophe where you lose customer records, payment history, or future bookings.

If your location is unsafe to operate

• Sign in to RentalTide from any device anywhere in the world.
• Mark affected dates as closed under Operations > Calendar > Block dates.
• Use the bulk messaging tools to email or text every customer with a booking in the affected window.
• Use the cart recovery tools to follow up once you reopen.

If you lose physical hardware (terminals, printers, tablets)

• Your data is fine.
• A new tablet or laptop with a browser is all you need to be operational again.
• Stripe will reissue terminals on request. Reach out and we will help you place the order.

If we lose office access

We have no on-premise servers, no local databases, and no critical infrastructure at our office. Engineering and support work from laptops that can be replaced overnight. The platform runs entirely on AWS and would not notice.

Offline procedures for staff

A quick reference your staff can use the moment internet drops, before they have a chance to read this whole page.

Print this table and laminate it. We have a printable PDF version on request.

How we communicate during incidents

You will hear from us in three places during a real incident:

1. Status page — status.rentaltide.com is updated within minutes of detection. This is the source of truth for current status, what is affected, and ETA.
2. Support channel — your dedicated support channel (Slack or email) gets a message within the same window with anything specific to your location.
3. Post-incident report — for any incident with customer impact lasting more than fifteen minutes, we publish a written postmortem within five business days. These are stored under Operations > Support channel and are also shared publicly when appropriate.

If you ever cannot reach us through your usual channel during an outage, the support email at support@rentaltide.com is monitored from outside our normal infrastructure.

Recovery time and recovery point at a glance

These are the targets we hold ourselves to. They are described in detail in the technical appendix.

Practising your own continuity plan

Reliability on our side is half the picture. We recommend every location:

• Keep paper walk-up sheets and waivers on hand for the season.
• Keep one staff phone with a cellular plan dedicated to RentalTide access during dock-wifi outages.
• Identify a backup person who can sign in and run the admin panel if the manager is unavailable.
• Test your Stripe terminal on cellular at least once per season.
• Know where to find your support channel and the status page.

If you would like help drafting a continuity plan specific to your location, reach out in your support channel and we will work through it with you.