Mendix Cloud Hosting on CloudFoundry

Editor's Notes

• #6: Full HA capable. Mendix has adopted the ISO27001:2013 framework. All controls are audited against an ISAE3402 Type II assurance standard, which is the International equivalent of a SOC1 Type II report. Annual test performed by an external auditor based on the 3 listed frameworks. Vulnerabilities impacting customer app will be communicated within 36 hours of detection. On request more info about the pen-test can be verbally provided, detailed findings are company confidential. 3th party pen test is allowed, only on the customer application
• #7: All platform components are hosted in the Mendix Cloud in EU (Frankfurt), the Runtime containers are hosted in the region of choice.
• #9: The primary hosting locations are as follows: Mendix Cloud EU: AWS Frankfurt Germany Mendix Cloud EU: AWS London UK Mendix Cloud US: AWS North Virginia USA Mendix Cloud Asia Pacific: AWS Tokyo Data will always be stored in the same political region for the following regions: Data in the EU, including backups, will stay within the EU Data in the US, including backups, will stay within the US Based on customer needs additional regions could be made available.
• #10: Each Mendix App can have 1 Mendix Cloud Node. A Mendix cloud node has a Production and Acceptance Environment. Optionally a third Environment can be added ‘Test’. Each environment takes an ‘App Container’, based on the purchased ‘App Container’ and cloud options a different setup is made available.
• #13: Each Mendix App can have 1 Mendix Cloud Node. A Mendix cloud node has a Production and Acceptance Environment. Optionally a third Environment can be added ‘Test’. Each environment takes an ‘App Container’, based on the purchased ‘App Container’ and cloud options a different setup is made available.
• #14: A standard App Container provides a Runtime container, an Amazon RDS database container and an S3 file storage bucket The size and capacity of these containers depends on the Cloud Edition that has been purchased (Strato, Meso, Iono, Magneto)
• #15: With Mendix 7.0 and higher Horizontal scaling is available Self-service through sliders, scaling can be configured within the contract parameters Runtime containers must be identical in size. Zero downtime for application failure (traffic gets automatically redirected to different runtime containers)
• #16: A standard App Container provides a Runtime container, an Amazon RDS database container and an S3 file storage bucket The size and capacity of these containers depends on the Cloud Edition that has been purchased (Strato, Meso, Iono, Magneto) With Mendix 7.0 and higher Horizontal scaling is available Self-service through sliders, scaling can be configured within the contract parameters Runtime containers must be identical in size. Zero downtime for application failure (traffic gets automatically redirected to different runtime containers)
• #17: CloudFoundry provides a health manager, this HM automatically detects if a container fails. In case of a container failure a the runtime container is replicated and started Replication of a container and starting an average Mendix application normally takes 1.5 minutes. A 60 second timeout + restarting the application (Larger projects with long startup time will take longer to recover) Next to the Mendix/CloudFoundry health manager, we also leverage the RDS health manager function. AWS detects a downtime Per environment a total SLA of 99.5% measured per month
• #18: The Mendix / CF infrastructure choses 1 AZ to run the environment in (equally dividing all environments across both AZs). This is to keep latency between db and runtime low. To guarantee data availability backups are stored in a different AZ or different Region.
• #19: Fallback environment provides a second RDS instance in the second AZ. In case of failure all communication will be redirected to fallback RDS. When RDS comes back online that RDS will serve as the Fallback AWS specifies that failover takes typically 60-120 seconds. http://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/Concepts.MultiAZ.html Multi-AZ runtime container functionality will remain the same Per environment a total SLA of 99.95% measured per month (max 20.2 minutes of downtime per month) S3 inherit architecture already facilitates multi-AZ hosting
• #20: CloudFoundry provides a health manager, this HM automatically detects if a container fails. In case of a container failure a the runtime container is replicated and started Replication of a container and starting an average Mendix application normally takes 1.5 minutes. A 60 second timeout + restarting the application (Larger projects with long startup time will take longer to recover) Per environment a total SLA of 99.5% measured per month Fallback environment provides a second RDS instance in the second AZ. In case of failure all communication will be redirected to fallback RDS. When RDS comes back online that RDS will serve as the Fallback AWS specifies that failover takes typically 60-120 seconds. http://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/Concepts.MultiAZ.html Multi-AZ runtime container functionality will remain the same Per environment a total SLA of 99.95% measured per month (max 20.2 minutes of downtime per month) S3 inherit architecture already facilitates multi-AZ hosting