Service Level Agreement:

This document shows you the high quality of service we are promising. RackNerd has spent considerable time to make available one of the most redundant networks available anywhere. This has been accomplished by the following:

- Redundant routers and switches

- Multiple upstream providers

- All key network components are monitored

This Service Level Agreement (“SLA”) applies to paid RackNerd shared hosting, reseller hosting, VPS, dedicated server, colocation, private cloud, and DRaaS services, unless a service order states otherwise. It covers network transit and electrical power provided by RackNerd. It does not guarantee that a client’s operating system, websites, applications, data, or backups will be available or recoverable.

1. Uptime Guarantee

RackNerd will provide 99.9% monthly uptime for (a) transit to the Internet and (b) electricity, measured separately for each affected service over each calendar month.

2. What “Unavailable” Means

“Unavailable” means that, due to a failure of RackNerd’s network or of electrical power supplied by RackNerd to the service:

- the service has no electrical power; or

- the service’s assigned public IPv4 address has no transit to or from the public Internet (for example, it cannot be reached from the public Internet, and cannot reach the public Internet).

Unavailability does not include any of the following:

- latency, jitter, routing changes, congestion, or degraded performance;

- partial packet loss;

- loss or poor performance to a specific destination, third-party network, or individual remote path, while the service remains reachable on the public Internet generally;

- the client’s operating system, firewall, software, configuration, virtual machines, websites, or applications being down, misconfigured, or offline;

- time needed by the client to reconfigure networking, rebuild bridges or interfaces, reboot virtual machines, restore applications, or otherwise recover the client’s environment after RackNerd has restored power and/or network transit to the service.

RackNerd determines unavailability using its own monitoring, logs, and records. Client evidence (such as MTR output) may be considered, but RackNerd’s determination is final.

3. Service-Specific Coverage

Shared hosting and reseller hosting: This SLA covers network transit to the shared server and electrical power to that server. It does not cover website or application errors, plugins, scripts, content management systems, email deliverability or spam filtering, DNS not controlled by RackNerd, resource-limit throttling, SSL/certificate issues caused by the client, or individual mailbox issues.

VPS: This SLA covers network transit to the VPS’s assigned public IPv4 address and electrical power to the host node. A host node or hypervisor failure that takes the VPS offline counts as unavailability, except to the extent excluded below. It does not cover the guest operating system, the client’s firewall or software, or the client’s own configuration.

Dedicated servers: This SLA covers network transit to the server’s assigned public IPv4 address and electrical power to the server.

Due to the nature of dedicated servers, and the fact that they are typically custom-built per customer specification, hardware needs can vary from server to server. RackNerd provides 24x7 support for dedicated hardware issues and, in practice, typically responds to failed hardware replacement requests within 4 hours or less. RackNerd also makes every effort to stock compatible replacement parts in every datacenter location.

Once a client requests replacement and RackNerd has confirmed that RackNerd-provided hardware has failed, RackNerd guarantees replacement of that hardware within 24 hours for purposes of this SLA, although in practice this is typically completed within a few hours.

If replacement of confirmed failed hardware takes longer than 24 hours, RackNerd will review the incident and issue a fair account credit. That credit will be at least 2% of the server’s Recurring Monthly Fee for each completed hour beyond the 24-hour replacement window, and will not exceed 50% of that fee for the month.

Network and power credits are handled under Sections 1 and 2. Dedicated hardware replacement is handled under this section. After replacement hardware has been installed, any time needed to reinstall or restore the operating system or data is not part of the 24-hour replacement window. Hardware supplied by the client, if any, is not covered.

Colocation: This SLA covers electrical power to the RackNerd-provided circuit or feed serving the colocation space, and network transit only if RackNerd provides that network service. Failure of client-owned or client-supplied equipment is not covered.

Private cloud and DRaaS: This SLA covers the infrastructure RackNerd provides (network transit and power to the applicable platform), on the same terms as VPS above. It does not cover the client’s virtual machines, operating systems, applications, replication jobs, or recovery point/recovery time objectives unless a separate written agreement says otherwise.

4. Credits

If network transit or electricity for an affected paid service is unavailable for a cumulative period beyond the 99.9% monthly uptime guarantee, the client will receive an account credit equal to 2% of that service’s Recurring Monthly Fee for each additional completed 1 hour of unavailability.

Credit rules:

- Credits apply only to covered unavailability beyond the 99.9% monthly guarantee.

- Only completed 1-hour increments are credited. Partial hours do not round up and do not count.

- “Recurring Monthly Fee” means the regular monthly recurring charge for the specific affected service only. If the service is billed quarterly, annually, or on another prepaid term, the Recurring Monthly Fee is the monthly equivalent of that service’s recurring charge (for example, the annual recurring charge divided by 12). Setup fees, add-ons that were not affected, overages, and other services on the account are not included.

- If dedicated hardware replacement takes longer than the 24-hour window in Section 3, RackNerd will issue a fair account credit of at least 2% of that service’s Recurring Monthly Fee for each completed hour beyond that window, up to 50% of that month’s Recurring Monthly Fee. Hardware credits and network/power credits are tracked separately, and together will not exceed 50% of that service’s Recurring Monthly Fee for the month.

- All credits for a service in a given month will not exceed 50% of that service’s Recurring Monthly Fee for the month in which the credit is issued.

- Credits are issued via the form of account credits only.

- Credits are the client’s sole and exclusive remedy for any failure to meet this SLA.

5. How to Request a Credit

To request a credit, the account owner must open a support ticket to our billing department (or via e-mail at [email protected]) within 30 days of the incident and include:

- the affected service (hostname, service ID, and/or IP address);

- the start and end time of the outage, in UTC; and

- a brief description of the issue.

Accounts must not be past due when requesting an SLA credit.

6. Maintenance

This SLA does not cover outages caused by scheduled or emergency network and/or facility maintenance.

- Scheduled maintenance will be announced in advance at https://status.racknerd.com/.

- Emergency maintenance may be performed without prior notice when reasonably necessary to protect the network, facility, hardware, or services. RackNerd will post notice at https://status.racknerd.com/ as soon as practicable.

- Combined scheduled and emergency maintenance will not exceed 180 minutes per calendar month for a given service. Maintenance time beyond that 180-minute cap may be treated as unavailability under this SLA, unless another exclusion applies.

An unexpected failure of RackNerd-provided power or network infrastructure (including a PDU, UPS, or switch) is not emergency maintenance. Downtime caused by that failure counts as unavailability under this SLA. Time spent repairing it is part of that same unavailability, unless it falls within a scheduled maintenance window announced in advance. For VPS services, an unexpected host node failure that takes the VPS offline is treated the same way, except as otherwise excluded in this SLA. For dedicated servers, hardware replacement is handled under Section 3.

RackNerd and its datacenter partners may use IPMI, iDRAC, iLO, KVM, or similar out-of-band management, and may reboot a service when reasonably necessary to diagnose, repair, or restore hardware, power, or network. When that work is scheduled or emergency maintenance, related downtime is excluded as described above.

7. Exclusions

This SLA does not cover, and no credit will be issued for, unavailability or degradation caused by any of the following:

- Equipment, software, configuration, or events under the client’s control, including the client’s operating system, firewall, applications, websites, virtual machines, scripts, plugins, and data;

- Dedicated server hardware, except for credits issued under Section 3 if confirmed replacement takes longer than 24 hours;

- Time required for the client to restore or reconfigure the client’s environment after RackNerd has restored power and/or network transit;

- Client-requested work, including reboots, reinstalls, rebuilds, migrations, IP changes, or other changes requested or approved by the client;

- Suspension, termination, rate-limiting, or nullrouting for non-payment, abuse, or violation of RackNerd’s terms of service or acceptable use policy;

- Denial-of-service attacks, or mitigation of such attacks, including temporary nullrouting or filtering;

- Issues affecting only a specific third-party destination, peer, ISP, or route, including blocking, filtering, or throttling by third parties or governments;

- Failures of third-party networks, DNS, or services not controlled by RackNerd;

- Backups, snapshots, backup jobs, or data recovery (these are not guaranteed by this SLA);

- Unavailability of IPMI, iDRAC, iLO, KVM, or other management interfaces as a standalone issue;

- IPv6-only issues if the service’s assigned IPv4 address remains reachable, or IPv4-only issues if the client’s service remains reachable over an assigned IPv6 address that RackNerd intended as a working path, as determined by RackNerd;

- Complimentary, promotional, trial, or other no-charge services;

- Services that are suspended, terminated, or pending cancellation;

- Force majeure events, including natural disasters, war, terrorism, labor disputes, government actions, pandemics, utility failures outside RackNerd’s facility, or cable cuts or other events outside RackNerd’s reasonable control.

8. General

This SLA may be updated from time to time. The version posted on RackNerd’s website at the time of an incident applies to that incident. This SLA does not modify RackNerd’s terms of service except as expressly stated here.

Last Revised 11.06.2026