Cloud Big Data Platform SLA
Rackspace Cloud Big Data Platform is a preconfigured group of virtual servers running Hadoop software (“Cluster”). Each Cluster is comprised of two related subsystems, the control plane and the data plane, each with its own service level guarantee.
I. CONTROL PLANE
The “control plane” is the Cloud Big Data API (Application Programming Interface) used to create, manage, and DELETE Clusters. Rackspace guarantees the control plane will be available 99.9% of the time in any given monthly billing period.
SLA Credit for Control Plane:
Monthly Availability Credit Percentage*
100% – 99.9% 0%
< 99.9% – 99.5% 10%
< 99.5% – 99.0% 20%
< 99.0% 30%
“Monthly Availability” is calculated on a per customer, per region basis, for a given monthly billing period, as follows (represented as a percentage):
1 – (Total API Errors)/(Total Valid API Requests)
* Credits will be calculated as a percentage of all Cluster fees in the affected region for the monthly billing period.
II. DATA PLANE
The “data plane” includes Clusters created via the control plane plus supporting systems and services required for the proper functioning and availability of the Clusters. Customers manage and have full control over their Clusters and Rackspace makes the following guarantees with regards to the following supporting systems and services:
Network
Rackspace guarantees that its data center network will be available 100% of the time in any given monthly billing period, excluding scheduled or emergency maintenance. Network downtime exists when a Cluster is not reachable as a result of a failure in the data center network.
SLA Credit for Network Downtime: 5% of the Cluster fees for each 30 minutes of network downtime (or portion thereof), up to 100% of the Cluster fees.**
Data Center Infrastructure
Rackspace guarantees that data center HVAC (Heating, Ventilation and Air Conditioning) and power will be functioning 100% of the time in any given monthly billing period, excluding scheduled or emergency maintenance. Data center infrastructure downtime exists when a Cluster experiences a failure as a result of power or heat problems.
SLA Credit for Data Center Infrastructure Downtime: 5% of the Cluster fees for each 30 minutes of data center infrastructure downtime (or portion thereof), up to 100% of the Cluster fees.**
Cluster Hosts
Rackspace guarantees the functioning of all Cluster hosts including the hypervisor. If a Cluster host fails, Rackspace guarantees that restoration or repair will be complete within one hour of problem identification.
SLA Credit for Cluster Host Downtime: 5% of the Cluster fees for each additional hour of downtime (or portion thereof), after the first hour, up to 100% of the Cluster fees. **
** SLA credits will be calculated as a percentage of the Cluster fees for those Clusters adversely affected by the failure for the monthly billing period during which the failure occurred.
III. DEFINITIONS AND LIMITATIONS
Definitions
For purposes of the Service Level Agreements outlined above:
• “API error” is defined as: (i) an HTTP 5xx server error response to a valid API request or (ii) no response to a valid API request because the API is down. Network errors or downtime outside of the Rackspace data center network do not constitute an API error;
• “Cluster” means a group of virtual servers preconfigured with Hadoop software;
• “Cluster fees” means the fees for your Cluster for the monthly billing period in which the failure occurred ;
• “cluster host” means the group of physical servers which hosts your Cluster;
• “data center network” means the portion of the Rackspace network extending from the network egress point of your Cluster to the outbound port of the data center border router;
• “emergency maintenance” means critical unforeseen maintenance (e.g. repairs, modifications, or upgrades) needed to ensure security or reliability;
• “power” excludes Cluster host power supplies which is covered as part of the Cluster hosts guarantee;
• “region” means a particular physical location where the Cluster exists (e.g. Dallas, Chicago, London). Cluster control and data planes are independent per region and do not span regions;
• “scheduled maintenance” means maintenance (e.g. repairs, modifications, or upgrades) that is announced at least 72 hours in advance;
• “valid API request” is defined as a well formed request that complies with the published API specification;
Limitations
• You are not entitled to a credit if you are in breach of your services agreement with Rackspace (including your payment obligations to Rackspace) until you have cured the breach. You are not entitled to a credit if downtime would not have occurred but for your breach of your agreement with Rackspace or your misuse of the Services. You are not entitled to a credit for downtime or outages resulting from denial of service attacks, virus activity, hacking attempts, or any other circumstances that are not within Rackspace’s control.
• To receive a credit, you must log a request via ticket in the Rackspace cloud control panel within thirty (30) days following the end of the downtime and provide logs evidencing Rackspace’s failure to meet the SLA. You must show that your use of your Cluster was adversely affected in some way as a result of the downtime to be eligible for the credit.
• Notwithstanding anything in this Service Level Agreement to the contrary, the maximum total credit for the monthly billing period, including all guaranties, shall not exceed 100% of your Cluster fee for that billing period. Credits that would be available but for this limitation will not be carried forward to future billing periods.
This Service Level Agreement is part of your Cloud Terms of Service.