SERVICE LEVEL AGREEMENT
SERVICE LEVEL AGREEMENT ADDENDUM VIRTUALISATION – VERSION 2014-05-28
This document is also available as a pdf file.
1. Definitions
1.1. Control Panel
Web interface that can execute maintenance tasks in the virtualisation layer.
1.2. CPU
The Central Processor Unit executes the instructions of the program to be processed.
1.3. CPU core
CPU's consist of multiple independent cores. The cores share a part of the physical architecture of the CPU.
1.4. Service
Virtual machine(s) or resource pool(s) on BIT's virtualisation platform.
1.5. I/O-operations
Reading and/or writing actions to and for the data used on the virtual machines.
1.6. Incident
When a service is no longer available due to causes other than maintenance in a maintenance interval, maintenance in consultation with the client, or misuse by the client.
1.7. Storage volume
Space on the hard disc available to the virtual machine(s) or resource pool(s).
1.8. RAM
Random Access Memory, volatile memory with random access.
1.9. Resource pool
The whole of virtual networks, storage, memory and virtual cores for the virtual machines.
1.10. Virtual Machine
A system in which to execute an operating system based on the available virtual resources. Including its own memory, CPU and storage.
2. SLA
This document is the addendum on the framework SLA for the service virtualisation.
3. Service Description
3.1. General
BIT manages a virtualisation platform. Through this platform, BIT provides clients with virtual machines and resource pools. This SLA is applicable to the availability of the resources for the virtual machines or the resource pool and the associated functionality in the control panel. Availability of the operating system or services on that operating system are not included in this SLA.
3.2. Availability resources
3.2.1. Redundancy
BIT reserves capacity on the physical infrastructure in several geographically separated data centers. Failure in one part of the physical infrastructure can thusly be compensated.
3.2.2. Memory
BIT guarantees the availability of the amount of RAM (memory) purchased by the client. These amounts are reserved.
3.2.3. CPU
The physical CPUs in the server systems are split into multiple virtual cores. BIT guarantees the availability of the amount of virtual CPU cores purchased by the client.
3.2.4 Control Panel
The control panel that is made available for maintenance of the virtual machines or the resource pool is hosted with load balancing from different data centers
3.3. Availability storage
3.3.1. Redundancy
The storage systems have been set up in a fully redundant setting, spread over two geographically separated data centers. The data that makes up the virtual machines and the resource pools are continually replicated between the separate storage systems. Failure of the systems in one of the data centers is compensated by the systems in the other data center.
3.3.2. Capacity
BIT reserves and guarantees the availability of the storage volume purchased by the client. These volumes are reserved.
3.3.3. Throughput
The available amount of bandwidth between the server and storage systems and the number of available I/O-operations are amply dimensioned. Peaks in the load can thusly be compensated.
3.4 Availability network
The availability of internet access of virtualisation services are described in the SLA addendum Internet Access.
4. Incidents
Incidents are classified into three categories by BIT:
| Priority | Description |
| 1 | The service is fully unavailable.
|
| 2 | The service is available but degraded.
|
| 3 | Incidents that barely cause the client any hindrance.
|
5. Non-Performance Penalties
In the event of non-observance with the defined availability, the client is entitled to compensation according to the following table:
| Priority | Time the service was unavailable | Non-performance penalty |
| 1 | 4 minutes, 19 seconds and more | 25% of the monthly sum |
| 1 & 2 | 2 hours and more | 50% of the monthly sum |
| 1 & 2 | 8 hours and more | 100% of the monthly sum |
The non-performance penalty is limited to 100% of the monthly sum of the service in question. The nonperformance penalty is limited to one penalty for a single incident case, even if the incident spreads over two calendar months or more. The penalty will only be rewarded upon customers request.