Regular platform maintenance and infrastructure improvements are scheduled in advance. Affected Customers are provided a start time and stop time for the applicable maintenance window and are notified at least forty-eight hours prior to beginning of the maintenance window. Acquia will proactively notify the Customer if maintenance will exceed a previously scheduled maintenance window or if maintenance needs to be postponed. Maintenance is scheduled from 11:00 PM (23:00) to 7:00 AM (07:00) local time for the following service regions:
In circumstances where security upgrades or other urgent maintenance must be performed, we undertake best efforts to notify Customers prior to beginning maintenance and to provide technical information as it is available. In these circumstances, maintenance windows begin immediately and continue until the emergency maintenance is completed. Customers are notified after this maintenance is complete and provided any additional instructions regarding required Customer actions. Acquia Global Support is available to clarify any questions or follow-up actions that are the result of the unscheduled maintenance.
Emergency hardware resizing
Acquia monitors traffic and capacity for Cloud Platform Enterprise subscriptions. If Acquia detects that your website has gone down or is in imminent danger of going down for any reason (such as a significant increase in web traffic or poor performing code), Acquia will add additional cloud capacity to your website. For applications running on Cloud Classic infrastructure, Acquia will contact you about upsizing your infrastructure, but will not wait for your prior approval. Customers will be responsible for any additional costs for the additional hardware, and any upsizing will remain in effect for a minimum of one week.
For applications running on the Cloud Next infrastructure, Cloud Platform capacity will increase as needed to ensure any activity in excess of typical patterns will not impact application performance or availability. Dynamic auto-scaling of environment capacity may be temporarily or permanently halted by Acquia if utilization so greatly exceeds expectations that a manual review by Acquia on-call personnel is required to validate activity and intervene in the event the surge in utilization is caused by unintentional application behaviors or some form of attack. Once this audit is complete, Acquia personnel may increase the auto-scaling limit on the affected environment to ensure additional capacity can continue to be allocated automatically as needed.
Customers on Cloud Next infrastructure may be contacted by Acquia to request a temporary or permanent increase in customer’s subscription tier if dynamically allocated additional capacity exceeds published Resource Limits guidelines for longer than three hours, or in a recurring fashion over the course of two or more weeks. If this pattern continues without a corresponding increase in subscription tier, Customer will be responsible for overage fees starting from the date of initial notification based on then-current applicable rates.
Acquia does not manage server capacity for Cloud Platform Professional subscriptions. Cloud Platform Professional is a self-service platform, and customers are responsible for monitoring their server capacity and resizing as needed. For more information, see Managing Cloud Platform infrastructure.
Infrastructure support
Acquia provides full infrastructure support to any Customer running its applications on Cloud Platform, Cloud Platform Enterprise, or Site Factory platforms. Due to Acquia’s platform configuration files containing Acquia’s intellectual property that is specific to our infrastructure, these files cannot be shared with Customers or partners.
Development and staging environments are designed and intended only for development and testing. Running production websites on development or staging environments is not supported, because they are not sized for production traffic and loads. In addition, development and staging environments may include shared resources which, if impacted by production-level traffic, may be in violation of terms of service. Acquia offers High-Availability and multi-region failover environments for customers with such requirements.
If a Customer chooses to host their application internally or with a third-party provider, Acquia Support is unable to provide troubleshooting or support assistance with the infrastructure. This limitation includes reviewing the Customer’s hardware, settings, and non-supported software infrastructure. Acquia Professional Services is able to provide assistance in these areas and with overall infrastructure architecture.
Installation of custom VCLs
Cloud Platform Enterprise and Site Factory customers with dedicated load balancers can request that custom VCL files be added to their load balancers. Customers that need custom VCL files must follow the instructions at Custom Varnish configuration. Custom VCL configurations are not supported with the Cloud Platform CDN.
Modification of the Varnish Configuration Language (VCL) Configuration
Cloud Platform provides a universal and optimized Varnish Configuration Language (VCL) configuration which cannot be customized on the shared platform. Customers with dedicated load balancers may provide a customized, tested VCL configuration for Acquia to implement on the Customer’s dedicated load balancers.
Acquia does not test customized VCL configurations. It is the Customer’s responsibility to test VCL changes in advance of supplying the configuration to Acquia. Prior to implementing the modifications on the Customer’s live balancer, Acquia will implement the VCL configuration changes on a load-testing balancer. It is the Customer’s responsibility to test these changes, confirm they are working as expected, and grant Acquia written permission to implement the changes on the live balancer. Acquia deploys custom VCL files on a set schedule so turnaround time to implement a custom VCL may be up to one week. Acquia may reject a modified VCL file if the modifications prove disruptive to the platform or generate high levels of platform alerts.
Cloud Platform Professional customers experiencing a significantly high traffic event will be contacted by Support and asked to divert traffic away from their site to prevent an outage for other customers sharing their infrastructure.
Customers using Cloud Platform Professional, Cloud Platform Enterprise, or Site Factory may utilize infrastructure shared with other customers. Sustained usage of any shared resource (such as, but not limited to: Cloud Next infrastructure, Cloud Classic staging servers, Cloud Classic load balancers, and version control servers) in excess of 20% of overall capacity, or otherwise to the extent that other customers are noticeably impacted, gives Acquia the discretion to move the customer to dedicated hardware at the customer’s expense, in order to preserve uptime for other customers also using the shared hardware.
Shared resources may include, but are not limited to:
Memory (RAM)
CPU
Disk
Network capacity
Customers with websites using the Acquia Search platform have limits for some or all of the following resources: total queries, documents, and disk space. If these limits are consistently met or exceeded, Acquia reserves the right to restrict the website’s use of the shared Search resources or move the customer’s website to a dedicated Acquia Search instance at the customer’s expense.
Removal from monitoring due to chronic application issues
If a website has issues with application stability and performance due to application issues which cannot be remedied without customer involvement, and as a result generates frequent alerts within Cloud Platform monitoring system, Acquia will notify the customer via a support ticket. This will identify and recommend measures to resolve or mitigate the issues causing the repeated alerts with an expectation that there is a mutual interest in stabilizing the website as soon as possible.
If a customer elects to take no action towards addressing the issues, and if there are three or more incidents with the same cause within 14 days, Acquia will temporarily remove the website from platform monitoring.
When a website is removed from monitoring, Acquia does not have visibility into the condition of your Cloud Platform instances and we cannot provide proactive alerts on certain issues impacting your application’s performance and availability.
Customers with applications removed from monitoring may request applications be added back to monitoring after demonstrating the underlying issues have been addressed. Normal support services will be provided to customers independent of website monitoring status.
Long-running queries
Database queries can be both generated by Drupal in response to website visitors displaying pages on which selective data is displayed, and through administrative or maintenance actions taken by a website developer or administrator.
Any query with a running time of greater than one second is a long-running query. A single particularly long-running query or a high volume of shorter long-running queries may negatively impact overall website performance, and can result in an interruption of a website’s availability. For recommendations about how to resolve slow queries, see the How to fix slow queries Acquia Knowledge Base page.
Terminating a task that contains a long-running query can result in data loss or a split-brain condition between databases in a High Availability environment. Due to these potential consequences, Acquia Support will not proactively terminate long-running queries without customer authorization.
In the event a long-running query is impairing website performance on Cloud Classic infrastructure, Acquia Support will attempt to contact the customer to obtain permission to terminate the query. Acquia Support will not terminate a query without confirmation from the customer and an acknowledgement of the potential consequences.
If Acquia Support cannot obtain confirmation from the customer, the query will be allowed to run. If the query impacts performance, website instances will be upsized as necessary to ensure stability.
On Cloud Next infrastructure, all slow queries lasting longer than 60 seconds on non-production environments and 5 minutes on production environments are terminated automatically to preserve application performance and prevent adverse impacts to pooled system resources.
Infrastructure, security & upgrades
Platform maintenance and security upgrades
Scheduled maintenance
Regular platform maintenance and infrastructure improvements are scheduled in advance. Affected Customers are provided a start time and stop time for the applicable maintenance window and are notified at least forty-eight hours prior to beginning of the maintenance window. Acquia will proactively notify the Customer if maintenance will exceed a previously scheduled maintenance window or if maintenance needs to be postponed. Maintenance is scheduled from 11:00 PM (23:00) to 7:00 AM (07:00) local time for the following service regions:
In circumstances where security upgrades or other urgent maintenance must be performed, we undertake best efforts to notify Customers prior to beginning maintenance and to provide technical information as it is available. In these circumstances, maintenance windows begin immediately and continue until the emergency maintenance is completed. Customers are notified after this maintenance is complete and provided any additional instructions regarding required Customer actions. Acquia Global Support is available to clarify any questions or follow-up actions that are the result of the unscheduled maintenance.
Emergency hardware resizing
Acquia monitors traffic and capacity for Cloud Platform Enterprise subscriptions. If Acquia detects that your website has gone down or is in imminent danger of going down for any reason (such as a significant increase in web traffic or poor performing code), Acquia will add additional cloud capacity to your website. For applications running on Cloud Classic infrastructure, Acquia will contact you about upsizing your infrastructure, but will not wait for your prior approval. Customers will be responsible for any additional costs for the additional hardware, and any upsizing will remain in effect for a minimum of one week.
For applications running on the Cloud Next infrastructure, Cloud Platform capacity will increase as needed to ensure any activity in excess of typical patterns will not impact application performance or availability. Dynamic auto-scaling of environment capacity may be temporarily or permanently halted by Acquia if utilization so greatly exceeds expectations that a manual review by Acquia on-call personnel is required to validate activity and intervene in the event the surge in utilization is caused by unintentional application behaviors or some form of attack. Once this audit is complete, Acquia personnel may increase the auto-scaling limit on the affected environment to ensure additional capacity can continue to be allocated automatically as needed.
Customers on Cloud Next infrastructure may be contacted by Acquia to request a temporary or permanent increase in customer’s subscription tier if dynamically allocated additional capacity exceeds published Resource Limits guidelines for longer than three hours, or in a recurring fashion over the course of two or more weeks. If this pattern continues without a corresponding increase in subscription tier, Customer will be responsible for overage fees starting from the date of initial notification based on then-current applicable rates.
Acquia does not manage server capacity for Cloud Platform Professional subscriptions. Cloud Platform Professional is a self-service platform, and customers are responsible for monitoring their server capacity and resizing as needed. For more information, see Managing Cloud Platform infrastructure.
Infrastructure support
Acquia provides full infrastructure support to any Customer running its applications on Cloud Platform, Cloud Platform Enterprise, or Site Factory platforms. Due to Acquia’s platform configuration files containing Acquia’s intellectual property that is specific to our infrastructure, these files cannot be shared with Customers or partners.
Development and staging environments are designed and intended only for development and testing. Running production websites on development or staging environments is not supported, because they are not sized for production traffic and loads. In addition, development and staging environments may include shared resources which, if impacted by production-level traffic, may be in violation of terms of service. Acquia offers High-Availability and multi-region failover environments for customers with such requirements.
If a Customer chooses to host their application internally or with a third-party provider, Acquia Support is unable to provide troubleshooting or support assistance with the infrastructure. This limitation includes reviewing the Customer’s hardware, settings, and non-supported software infrastructure. Acquia Professional Services is able to provide assistance in these areas and with overall infrastructure architecture.
Installation of custom VCLs
Cloud Platform Enterprise and Site Factory customers with dedicated load balancers can request that custom VCL files be added to their load balancers. Customers that need custom VCL files must follow the instructions at Custom Varnish configuration. Custom VCL configurations are not supported with the Cloud Platform CDN.
Modification of the Varnish Configuration Language (VCL) Configuration
Cloud Platform provides a universal and optimized Varnish Configuration Language (VCL) configuration which cannot be customized on the shared platform. Customers with dedicated load balancers may provide a customized, tested VCL configuration for Acquia to implement on the Customer’s dedicated load balancers.
Acquia does not test customized VCL configurations. It is the Customer’s responsibility to test VCL changes in advance of supplying the configuration to Acquia. Prior to implementing the modifications on the Customer’s live balancer, Acquia will implement the VCL configuration changes on a load-testing balancer. It is the Customer’s responsibility to test these changes, confirm they are working as expected, and grant Acquia written permission to implement the changes on the live balancer. Acquia deploys custom VCL files on a set schedule so turnaround time to implement a custom VCL may be up to one week. Acquia may reject a modified VCL file if the modifications prove disruptive to the platform or generate high levels of platform alerts.
Cloud Platform Professional customers experiencing a significantly high traffic event will be contacted by Support and asked to divert traffic away from their site to prevent an outage for other customers sharing their infrastructure.
Customers using Cloud Platform Professional, Cloud Platform Enterprise, or Site Factory may utilize infrastructure shared with other customers. Sustained usage of any shared resource (such as, but not limited to: Cloud Next infrastructure, Cloud Classic staging servers, Cloud Classic load balancers, and version control servers) in excess of 20% of overall capacity, or otherwise to the extent that other customers are noticeably impacted, gives Acquia the discretion to move the customer to dedicated hardware at the customer’s expense, in order to preserve uptime for other customers also using the shared hardware.
Shared resources may include, but are not limited to:
Memory (RAM)
CPU
Disk
Network capacity
Customers with websites using the Acquia Search platform have limits for some or all of the following resources: total queries, documents, and disk space. If these limits are consistently met or exceeded, Acquia reserves the right to restrict the website’s use of the shared Search resources or move the customer’s website to a dedicated Acquia Search instance at the customer’s expense.
Removal from monitoring due to chronic application issues
If a website has issues with application stability and performance due to application issues which cannot be remedied without customer involvement, and as a result generates frequent alerts within Cloud Platform monitoring system, Acquia will notify the customer via a support ticket. This will identify and recommend measures to resolve or mitigate the issues causing the repeated alerts with an expectation that there is a mutual interest in stabilizing the website as soon as possible.
If a customer elects to take no action towards addressing the issues, and if there are three or more incidents with the same cause within 14 days, Acquia will temporarily remove the website from platform monitoring.
When a website is removed from monitoring, Acquia does not have visibility into the condition of your Cloud Platform instances and we cannot provide proactive alerts on certain issues impacting your application’s performance and availability.
Customers with applications removed from monitoring may request applications be added back to monitoring after demonstrating the underlying issues have been addressed. Normal support services will be provided to customers independent of website monitoring status.
Long-running queries
Database queries can be both generated by Drupal in response to website visitors displaying pages on which selective data is displayed, and through administrative or maintenance actions taken by a website developer or administrator.
Any query with a running time of greater than one second is a long-running query. A single particularly long-running query or a high volume of shorter long-running queries may negatively impact overall website performance, and can result in an interruption of a website’s availability. For recommendations about how to resolve slow queries, see the How to fix slow queries Acquia Knowledge Base page.
Terminating a task that contains a long-running query can result in data loss or a split-brain condition between databases in a High Availability environment. Due to these potential consequences, Acquia Support will not proactively terminate long-running queries without customer authorization.
In the event a long-running query is impairing website performance on Cloud Classic infrastructure, Acquia Support will attempt to contact the customer to obtain permission to terminate the query. Acquia Support will not terminate a query without confirmation from the customer and an acknowledgement of the potential consequences.
If Acquia Support cannot obtain confirmation from the customer, the query will be allowed to run. If the query impacts performance, website instances will be upsized as necessary to ensure stability.
On Cloud Next infrastructure, all slow queries lasting longer than 60 seconds on non-production environments and 5 minutes on production environments are terminated automatically to preserve application performance and prevent adverse impacts to pooled system resources.
Did not find what you were looking for?
If this content did not answer your questions, try searching or contacting our support team for further assistance.
