CLOUD DATABASE SLA

On this page you will find all needed information on our SLA regarding our Cloud Database service

Last update: May 11th 2022

Article 1. Definition of terms

 

Term 

Definition  

Cloud 

Customer-based infrastructure as supplied by Tilaa 

Cloud Database 

Database service hosted in the Tilaa Cloud environment.  

Customer 

A person or organization to which paid services are delivered. 

Downtime 

The period the availability of the service is below 99% of the time, measured for the timeframe of one month.  

Error Rate 

The percentage of unsuccessful connections with valid requests reaching the service. 

General terms and conditions 

The description of the general conditions and regulations that apply to all products and services as delivered by Tilaa. 

Highly Available  

The database service is spread out across multiple VPS systems, having one primary database configuration and one or more redundant systems for synchronisation. 

Managed service 

The part of the cloud database configuration that is set-up and maintained by Tilaa.  

Scheduled maintenance 

The Downtime period related to network, hardware, or software maintenance or upgrades, with customer notification. 

RPO 

Recovery Point Objective, being the nearest point in time where the data is available for restore. 

RTO 

Recovery Time Objective, being the time between the failover and the database being fully available. 

Single Node  

The database service runs on a single VPS.  

Service 

A Service (or Services) or Ancillary Service[s] to be supplied by Tilaa based on the Agreement and Service[s] or Ancillary Service[s] that may be supplied on a fee-paying basis. 

SLO 

Service Level Objective, being the key performance indicator for the specific part of the service, such as guaranteed availability.  

Three node option 

A setup with three separate database nodes, which are asynchronously replicated. 

Tilaa 

Tilaa B.V., established in Amsterdam and registered with the Amsterdam Chamber of Commerce and Industry under file number 59712309. 

Two node option 

A setup with two separate database nodes, which are asynchronously replicated. 

Valid request 

The requests, issued to comply with the documentation and resulting in a non-error response. 

VPS  

Virtual Private Server within the Tilaa Cloud environment. 

Article 2. Preface

In spring 2022, Tilaa launched a new service providing a cloud-based database environment, the Tilaa Cloud Database. This addition to the existing portfolio provides users with a managed infrastructure that holds a database engine. This agreement describes the levels and actions of service applicable to the Tilaa Cloud Database environment. The agreements in this document are meant to be additional to the general terms and conditions, as stated on the Tilaa website. In situations where the definition in this document differs from the statement in the general terms and conditions, the SLA document prevails. 

 

Article 3. Service Definition 

  1. The Tilaa Cloud Database service offers a database engine within the datacentre-environment of Tilaa, with all the known advantages this holds. The service offers subscription-based database functionality with regards to database services. The current implementation holds MySQL engines, combined with scalable storage and functional monitoring.  

    The database service comes in three different layers with respect to the availability, being: 

    1. Single node service 
    2. A redundant version with two nodes 
    3. Highly available service with three nodes (primary and two replication sets). 

    Regarding the redundancy, with the single node option there is, hence the name, no redundancy at all. With the 2-node option, a second database engine provides single redundancy of the configuration. The HA-option has a double redundancy across three nodes, with a downtime of nearly zero (0,001% calculated, ~0% in practice).  

    Within the Highly Available configuration, each write operation to the primary database engine is replicated across two other nodes. In the event the primary node is unavailable, one of the remaining nodes will be declared active, taking over all write- and synchronization actions as the new primary node.  

    Despite the fact that Tilaa is managing the infrastructure of the database cluster environment involved in the use and management of the implementation of the database service, it will not perform database administration activities. This will be the sole responsibility of the customer. Tilaa will, however, provide the customer with the necessary data and tooling to obtain a stable, highly available database engine. In other words: the administration and/or management of the database infrastructure (nodes, storage, network, etc.) is done by Tilaa. The customer has its own responsibility for the definition of the database model, the implementation of the model within the engine and the use of the database environment. As the database service is still within the customers’ shared environment, the general fair-use policies regarding egress data and storage IOPS apply, as stated in Tilaa’s general terms and conditions.   

    As an addition to all three options, a backup service is available. The RPO depends on the number of backups that are made within a daily timeframe, set as a customer requirement. The RTO depends on the size of the database. Restores are done into a separate database within the cloud database service, considering the maximum throughput of the customers’ subscription. Since Tilaa does not perform database administration tasks, functional restores of the database are on best-effort basis. The service offers various levels of retention, with a standard period of one week (7 days). 

    As part of the infrastructure, Tilaa offers a highly configurable monitoring system that enables the customer to actively watch the environments behaviour with a standard set of metrics and add other custom ones to the monitoring engine.   

    Upon requesting and implementing a Cloud Database service, the customer receives the information needed for adoption and usage. This contains: 

    • The endpoint of the database service 
    • The root (admin) account of the database environment 
    • Entrypoint and credentials for the monitoring environment 

Article 4. Service fees

The Tilaa Cloud Database Service is available in three different options: 

  1. The single-node option, without redundancy 
  2. The two-node option, with single redundancy.
  3. The multi-node (3) option with dual redundancy and high availability 

Within these options, several configurations are made available, based on the size of the underlying VPSs. These are: 

Single-node option 

Redundant option 

3-node (HA) option 

Name 

Configuration 

(# nodes; Mem GB; cores; Storage GB) 

Name 

Configuration 

(# nodes; Mem GB; cores; Storage GB) 

Name 

Configuration 

(# nodes; Mem GB; cores; Storage GB) 

Level 1 

1; 2; 1; 50 

 

 

 

 

Level 2 

1; 4; 2; 100 

Level 2 

2; 4; 2; 100 

Level 2 

3; 4; 2; 100 

Pro 1

1; 8; 4; 150 

Pro 1 

2; 8; 4; 150 

Pro 1 

3; 8; 4; 150 

Pro 2 

1; 16; 6; 200 

Pro 2 

2; 16; 6; 200 

Pro 2

3; 16; 6; 200 

Pro 3 

1; 32; 8; 400 

Pro 3 

2; 32; 8; 400 

Pro 3 

3; 32; 8; 400 

Pro 4 

1; 16; 64; 400 

Pro 4 

2; 16; 64; 400 

Pro 4 

3; 16; 64; 400 

 

  1.  

 

Article 5. Service Levels

The service levels for the Cloud Database Service are divided into two parts, based on the implemented option.   

In addition to the general terms and conditions, the Cloud Database Service contains some extras, mainly addressing the highly available option. 

For the single node option as well as the redundant option, the SLA defines a minimum uptime of 99.9%. With the option being solely dependent on one VPS, there are different options and terms for uptime and service. The availability is defined by the uptime of the VPS, with restores are restore-times being defined by the size of the database and the assigned VPS resources. Restore of a single database onto the Cloud Database environment is possible but can have a major unpredictable impact on the availability and performance of the service.  

Since there is no possibility to swap the master nodes of the database service, it will be more vulnerable to disruptive actions by the customer. Moreover, the scheduled maintenance is expected to be longer and more frequent than the highly available option.  

This option comes with a single endpoint, directly attached to the VPS, without the means to balance the connections over more than one node.  

 The highly available option has been built with a minimum uptime of 99.99%. With this option comes the ability to swap the master database node without loss of service and/or data. From the master node, database actions are asynchronously replicated on the other two nodes, with the ability to switch over (swap) to another node at any time, making it the new master node. The databases, and corresponding files, are stored on three local storage volumes of the three individual nodes. In case the master node is swapped, the new master will use its designated storage. The failover is done automatically. 

The endpoint of the database service, being a single URL, is ‘balanced’ to the master node. When a master-swap is issued, for instance when the master has become unavailable, the endpoint is set to use the new master. 

The default SLA times are divided into two major groups. 

  • Standard: 5 x 8 service  
  • Gold: 24 x7 service 

Assistance and support on the Tilaa Platform during the helpdesk opening hours, five [5] days a week from Monday to Friday 09.00 to 18.00.  

Access to Tilaa’s 24-hour emergency telephone support is only available in the event of a layer 1 failure.  

Assistance and support on the OS of up to half an [0.5] hour a month with a first response time of one [1] hour on weekdays from 09.00 to 17.00 (Dutch time zone) excluding Sundays and public holidays generally observed in The Netherlands. 

 

Article 6. SLA infringement penalties 

In the event an SLO cannot be met on the 3-node highly available option, this will impact the invoice of the service in the next period according to the following table:  

Monthly Uptime Percentage 

Percentage of monthly bill for Covered Service which does not meet SLO that in Tilaa Credits, to be used within the timeframe of one month. 

99.0% - < 99.95%  

10%  

95.0% - < 99.0%  

25%  

< 95.0%  

100%  

 

The service has a nominal uptime SLO of 99.99%, not considering the downtime with respect to the scheduled maintenance or security updates on the service. 

 

Article 7. Access to data

As Tilaa will be the supplier of the infrastructure, including the database engine, it also will issue a master (root) account for the database and designated VPS systems. Tilaa will and shall not apply this means of access without written approval of the customer.  

It will be possible for Tilaa to hand over logfiles of the VPS and the database engine for customer review and/or analysis.  

    1.  

 

Article 8. Communication

As with all Tilaa services, the main means of communication is through the support channel via E-mail or phone. Every registered customer can issue a ticket holding their questions and disruptive issues in the service management portal. Tilaa support will react on this with the standard support level, which is defined based on the severeness of the issue. 

If in a specific case the customer is unsatisfied with the support given, an escalation level is in place through the operational manager.  

 

Article 9. Exclusions

The general terms and conditions of Tilaa, as found on the website, also apply to this Cloud Database service. Additional exclusions are: 

As stated in the service definition, Tilaa is involved with maintaining the availability of the service, and its functional components from a technical perspective. Implementing the Tilaa Cloud Database service into any solution will be the main responsibility of the customer.  

  • Tilaa does not offer DBA services, nor development of applications using the Cloud Database Service. This includes database provisioning or functional database-management. 
    If any action originating or initiated from the customer-side are causing downtime or have any other negative effect on the availability of the service, no infringement penalties are applied towards Tilaa.  
  • The supported database engines are fully based on open-source availability of the software. If a specific version is regarded “End Of Service”, Tilaa will not be able to offer this version any longer. 
  • The Tilaa Cloud database is based on a configuration of our standard VPS offerings, in conjunction with the local hypervisor storage. This will give guaranteed performances. By using other storage-environments, like the Tilaa BigDisk object storage, the performance of the database is not specified and reliable.