On this page
1. Overview
This Service Level Agreement ("SLA") is part of the FlexDedi Terms of Service and applies to every Server. It sets two commitments:
- Network availability: 99.99% monthly availability of each Server's public network port
- Hardware replacement: failed hardware components replaced within 4 hours of diagnosis by our engineers
If we miss either commitment, you can claim service credits as described below. Capitalized terms not defined here, such as Server and Services, have the meanings given in the Terms.
2. Definitions
- Month: a calendar month, measured in UTC.
- Monthly fee: the monthly price of the affected Server, excluding taxes.
- Downtime: a period during which a Server's public network port cannot send or receive traffic to and from the internet because of a failure in FlexDedi's network, as confirmed by our monitoring or by a support ticket.
- Monthly availability: the share of the Month without Downtime, calculated as shown below.
- Scheduled maintenance: maintenance announced to affected customers at least 48 hours in advance.
- Diagnosis: the moment our engineers confirm that a hardware component has failed.
Monthly availability is calculated for each Server as follows:
Monthly availability (%) = (minutes in the Month - minutes of Downtime) / minutes in the Month × 100A 30-day Month has 43,200 minutes, so 99.99% availability allows about 4.3 minutes of Downtime.
3. Network availability commitment
We commit to 99.99% Monthly availability for the public network port of each Server. Availability is calculated separately for each Server, and Downtime on one Server does not count against another.
Downtime starts at the earlier of the moment our monitoring detects the loss of connectivity and the moment you open a support ticket reporting it, provided we confirm that the cause lies in our network. It ends when connectivity is restored.
For Servers with two ports (2 × 25 or 2 × 100 Gbps), Downtime occurs only when the Server has no public connectivity through either port. Reduced capacity while one link is down is not Downtime, but please report it so we can fix it.
This SLA covers connectivity, not throughput, latency or packet loss. It does not cover the control panel, API, out-of-band console, private networking or delivery times. You can follow the state of our network on the status page.
4. Exclusions
The following are not counted as Downtime:
- scheduled maintenance announced at least 48 hours in advance
- issues caused by you, your users or your customers, including network or firewall misconfiguration, and reboots, reinstalls, rescue mode or power actions started through the client area or API
- problems with the operating system or software on the Server, including images installed from our OS library
- actions taken by our DDoS protection to mitigate an attack on your Server, such as filtering, rate limiting or temporarily null-routing the targeted IP address
- Services suspended under the Terms of Service or the Acceptable Use Policy
- force majeure events, as defined in the Terms of Service
- failures of hardware inside your Server, which are covered by the hardware replacement commitment in section 5
- problems on networks outside FlexDedi's network, such as a visitor's internet provider
5. Hardware replacement commitment
We keep spare parts on site in every location. When a hardware component fails, such as a drive, power supply, memory module or network card, our engineers replace it within 4 hours of Diagnosis. They start diagnosing as soon as you report the fault through support or our monitoring detects it.
If a fault cannot be fixed by replacing a single component, we may move your Server to equivalent or better hardware, transferring your drives where possible so your data is kept. The same 4-hour commitment applies.
When a replacement requires powering off the Server, we coordinate the intervention with you. Time spent waiting for your approval, for information we need from you, or for a time slot you request does not count toward the 4 hours.
The commitment covers the hardware only. Because Servers are unmanaged, rebuilding RAID arrays, reinstalling software and restoring data remain your responsibility. Our software RAID guide explains how to rebuild an mdadm array after a drive replacement.
6. Service credits
Network availability credits
| Monthly availability | Downtime in a 30-day Month (approx.) | Credit |
|---|---|---|
| 99.99% or higher | Up to 4.3 minutes | None, the commitment is met |
| Below 99.99%, at least 99.9% | 4.3 to 43.2 minutes | 5% of the Monthly fee |
| Below 99.9%, at least 99.5% | 43.2 minutes to 3.6 hours | 10% of the Monthly fee |
| Below 99.5%, at least 99.0% | 3.6 to 7.2 hours | 25% of the Monthly fee |
| Below 99.0% | More than 7.2 hours | 50% of the Monthly fee |
For example, if your Pro 32 ($749/mo) has 50 minutes of Downtime in a 30-day Month, its Monthly availability is (43,200 - 50) / 43,200 × 100 = 99.88%. That is below 99.9% but at least 99.5%, so the credit is 10% of $749, or $74.90.
Hardware replacement credits
If a failed component is replaced more than 4 hours after Diagnosis, you receive a credit of 5% of the affected Server's Monthly fee for each additional started hour, up to 50% of the Monthly fee per incident.
For example, if a failed NVMe drive in your Core 16 ($299/mo) is diagnosed at 10:00 UTC and replaced at 15:20 UTC, the replacement took 1 hour 20 minutes longer than the 4-hour commitment. That counts as 2 started hours, so the credit is 10% of $299, or $29.90.
Limits
Total credits for a Server in a Month, for network availability and hardware replacement combined, are capped at 100% of that Server's Monthly fee. Credits are applied to future invoices, have no cash value and cannot be transferred to another account.
7. How to claim credits
Open a claim through support within 30 days of the incident, with:
- the affected Server's identifier and main IP address
- the start and end time of each incident, with time zone
- a short description of the issue and any related ticket numbers
We check each claim against our monitoring and ticket records and reply with our decision. Approved credits appear on your next invoice. Claims made after 30 days, or without the information above, are not eligible.
8. Sole remedy
Service credits are your sole and exclusive remedy for Downtime and for delays in hardware replacement, except where the law provides otherwise. They do not affect your right to cancel a Service at any time under the Terms of Service.
9. Changes to this SLA
We may update this SLA as described in the Terms of Service, with at least 30 days' notice of material changes. A change never reduces credits for incidents that happened before it takes effect.