Availability
Uptime SLA
Progress Software’s goal for the production Sitefinity websites hosted in Sitefinity Cloud is to be available 99.9% of the time in any given monthly billing period, and Progress Software exercises commercially reasonable efforts toward this goal. Monthly availability of the production Sitefinity websites hosted in Sitefinity Cloud is calculated by subtracting from 1 the total minutes, in which the production Sitefinity website hosted in Sitefinity Cloud has returned HTTP status code 5XX
in response to a web request to this website (“Downtime Minutes”) for the billing month divided by 43829.1 (the average number of minutes in any given month).
-
Incident notifications over email - send an automated email to a set of customer defined email addresses that notifies when an incident has been detected/resolved.
-
Incident notifications for customer owned PagerDuty accounts - in addition to the existing functionality to send emails to customers when there is an availability incident, we have added support for notifications to customer owned PagerDuty accounts.
Health monitoring
The health of the application is monitored by health checks that run every minute from 5 different locations around the globe. If health checks fail for 2 or more locations within a 5 minute interval, an incident is automatically triggered and the Sitefinity Cloud On-Call team is engaged to resolve the issue.
For more information, see Track website availability (uptime).
Disabling health monitoring due to chronic application issues
As result of a root cause analysis (RCA), it might be determined that the issue(s) impacting the application's stability and performance cannot be remedied without customer involvement. In such cases, the information on the required measures to troubleshoot, resolve, or mitigate the issue(s) will be provided to the customer via the support case that is opened for the availability incident(s).
If a customer decides to take no action towards addressing the issues, and if there are three or more incidents with the same root cause within 14 days, Progress Software will temporarily disable the health monitoring for the application.
When health monitoring is disabled, Progress Software does not have visibility into the availability of the application and will not provide proactive availability management.
The affected customer can request for health monitoring to be enabled after demonstrating the underlying issues have been addressed. Normal support services will be provided to the customer independent of health monitoring status.
Go-Live of your Sitefinity Cloud application
Health monitoring is enabled after the Go-Live with one or more website(s) that are hosted on the application. Before the Go-Live, health monitoring is disabled to avoid false-positive alerts during the development phase, when frequent changes and not fully-tested code can be deployed to Production.
For more information related to the Go-Live, see Prepare for Go-Live.
On-Duty team
A team of specially trained engineers that is on-call 24x7 and can respond to availability and security incidents within minutes to diagnose and resolve the issues, as well as provide RCA and steps for future prevention, according to the required SLA.
Zero-downtime deployments
Sitefinity Cloud leverages Azure DevOps for CI/CD automation, with source control and versioning, prebuilt pipelines for code promotion, staging and production environments, a deployment dashboard and comprehensive logs for troubleshooting and detailed analysis. The build pipeline has a tight approval process and failsafe features for zero-downtime deployments.
Multi-region failover
Multi-region failover provides a robust disaster recovery option, ensuring against critical incidents and infrastructure outage, and enabling the Sitefinity Cloud on-call team to restore service availability within up to 30 min. in case of regional disaster with the Azure data center.
For more information about this feature, see Multi-region failover.
Automated Azure Redis cache memory management
Redis cache memory management is automated to handle memory leak cases and purge the cache when memory exceeds 90% for prolonged periods of time.