Redis Cloud Service Level Agreement (SLA)

1. Definitions.

Active-Active SLA applies to Active-Active Database deployments.

Database means a Redis database instance deployed by Customer within Redis Cloud.

Deployment Minutes means the total number of minutes that a Database has been deployed with the Cloud Services during a billing month.

Downtime means the total accumulated Deployment Minutes (across all Databases deployed by a Customer under a Commercial Subscription) during which the Database is unavailable. A minute is considered unavailable for a Database if repeated attempts by a Customer to establish a connection to the Database within that minute fail when tested at intervals of no more than one minute. For Active-Active SLAs, a minute is considered unavailable if repeated attempts by a Customer to establish a connection to all the Databases under Active-Active deployment within that minute fail when tested at intervals of no more than one minute. "Downtime" does not include (a) downtime for Maintenance or (b) partial minutes of unavailability. For RDI, “Downtime” means the total accumulated minutes during which the RDI Management API made available to the Customer is unavailable. A minute is considered unavailable for RDI if during that minute, all continuous attempts by Customer to access such RDI Management API fail.

Maintenance means the incorporation of new features, upgrades, updates, cluster optimization, patches, and/or bug fixes performed on the Cloud Services as described at https://redis.io/docs/latest/operate/rc/subscriptions/maintenance/.

Maximum Available Minutes means the sum of all Deployment Minutes across all Databases deployed by a Customer in a Cloud Services subscription during a billing month. For RDI, “Maximum Available Minutes” means the sum of minutes in a billing month during which the RDI Management API is expected to be available under the applicable Customer’s Cloud Services Commercial Subscription.

Monthly Uptime Percentage for a Database is calculated as the (Maximum Available Minutes - Downtime) / Maximum Available Minutes in a billing month for a Cloud Services subscription. For RDI, “Monthly Uptime Percentage” is calculated as the (Maximum Available Minutes for the RDI Management API - Downtime for RDI) / Maximum Available Minutes of RDI in a billing month for a Workspace.

Multi-AZ SLA applies to multi-availability zone deployments that are not part of an Active-Active deployment.

Pipeline means an RDI data flow configured within a Workspace to capture, process, and deliver RDI Data from a Source Database to a Redis Target Database.

RDI means Redis Data Integration managed as Cloud Services used to configure, manage, monitor, and operate RDI data integration workflows through the RDI Management API and the Customer’s interface.

RDI Data means the information stored in the Source Database by Customer that RDI writes, deletes, or updates to the Redis Target Database.

RDI Management API means a control-plane API for RDI that Customers use to configure, manage, monitor, and operate Workspaces and Pipelines.

Redis Target Database means the Redis Database to which RDI writes, deletes, or updates RDI Data.

Service Credit is a credit toward future Cloud Services calculated in accordance with Section 3.3, and is Customer's sole and exclusive remedy for any failure by Redis to meet the Service Commitment.

Source Database means the source Customer database, system, or service from which RDI captures, reads, or ingests data for delivery to the Redis Target Database.

Standard SLA applies to Database deployments that are not part of an Active-Active deployment and have replication enabled, or a Multi-AZ deployment. This is the only SLA level applicable to Essentials plans.

Workspace means the RDI environment, provisioned as part of the Cloud Services, in which Customer configures, manages, and operates RDI resources and related RDI Data integration functionality.

2. Plans.

Redis provides the Cloud Services in three paid plans: (i) Essentials, (ii) Pro, and (iii) Annual. The applicable plan determines the Customer's fees, Transaction method, Maintenance Window permissions, Monthly Uptime Percentage, and infrastructure configuration, as described at https://redis.io/redis-enterprise-cloud/pricing/.

3. Service Levels and Remedies.

3.1 Service Commitments. The Service Commitments and Service Credits apply only to Cloud Service deployments with replication enabled that have been up for a minimum of 24 hours, and apply separately to each Commercial Subscription. Redis will use commercially reasonable efforts to provide Monthly Uptime Percentages of: (a) at least 99.999% if the Active-Active SLA applies; (b) at least 99.99% if the Multi-AZ SLA applies; or (c) at least 99.9% if the Standard SLA applies (the applicable "Service Commitment"). If the Cloud Services do not meet the applicable Service Commitment, Customer will be eligible to receive a Service Credit as described below. For clarity, single-instance deployments without replication enabled are not covered by this SLA.

3.2 RDI Service Commitments. Redis will use commercially reasonable efforts to provide a Monthly Uptime Percentage of at least 99.9% (the “RDI Service Commitment”). In the event the RDI Management API does not meet this RDI Service Commitment, Customer will be eligible to receive a Service Credit as described below.

3.3 Service Credits. Service Credits are calculated as a percentage of the monthly charges paid by Customer for the affected Commercial Subscription inaccordance with the following table:

Table 1 – Service Credit Calculations:

Service Credits will be applied against future payments for Cloud Services or RDI only. Service Credits do not entitle Customer to any refund or payment from Redis and may not be transferred or applied to any other account. In no event will the aggregate Service Credits for any billing month exceed 25% of the monthly charges for the affected Commercial Subscription in that billing month. Customer's sole and exclusive remedy for any unavailability, non-performance, or other failure by Redis to provide Cloud Services or RDI in compliance with the Service Commitment or the RDI Service Commitment, including any claim relating to the availability, uptime, or performance of the Cloud Services or RDI, is the receipt of Service Credits in accordance with this SLA. For the avoidance of doubt, any claims relating to the availability, uptime, or performance of the Cloud Services or RDI, including any warranty claims under the Agreement, are governed exclusively by this SLA, and the remedies in this Section are the sole remedies for such claims.

3.4 Requesting Service Credits. To receive a Service Credit, Customer must: (a) log a support ticket with Redis within twenty-four hours of first becoming aware of the Downtime event; and (b) submit a written request to Redis by the end of the second billing cycle after the incident, including: (i) "SLA Credit Request" in the subject line; (ii) the dates and times of each Downtime incident; (iii) the affected Database name(s), cloud name(s), Workspace location(s), and region(s); (iv) logs documenting the errors (with confidential information redacted); and (v) a description of Customer's attempts to resolve the Downtime. Customer must reasonably assist Redis with problem diagnosis and resolution. If Redis confirms, based on Redis’ service monitoring, logs, or other relevant information, and taking into account Customer’s supporting evidence, the Monthly Uptime Percentage fell below the Service Commitment or RDI Service Commitment, Redis will issue the Service Credit within one billing cycle following receipt of the request. Failure to submit a compliant request disqualifies Customer from receiving Service Credits. Customer must be in compliance with the Agreement to be eligible for a Service Credit.

4. Exclusions.

4.1 General Exclusions. This SLA does not apply to any performance or availability issues (and such will not count toward any Downtime calculation) resulting from:

(a) suspension or termination of Cloud Services under the Agreement, or Preview or Evaluation Services deployments;

(b) Customer's or any third party's actions or inactions, including misconfigurations, failure to follow the Documentation or appropriate security practices, faulty input, unsupported operations, or unauthorized access to or use of the Cloud Services;

(c) factors outside Redis' reasonable control, including Force Majeure Events, network or device failures at Customer's site or between Customer's site and the Cloud Services, and failures of third-party services, hardware, software, or bandwidth;

(d) Customer exceeding the Provisioned Workload after Redis has notified Customer that additional resources are required, or Customer's continued use of the Cloud Services contrary to Redis' written recommendations;

(e) Maintenance, or expiration of a Grace Period without resolution due to Customer's lack of cooperation, or use of outdated versions of the Cloud Services beyond the support periods in the Support Policy;

(f) Customer's attempts to perform operations that exceed prescribed quotas, or Redis' throttling of suspected abusive behavior;

(g) under BYOC deployments, Customer's failure to maintain the access permissions required under Appendix 1 of the Agreement, for the duration of such failure; or

(h) unsupported platforms, configurations, or third-party modifications.

If availability is impacted by factors other than those used in the Monthly Uptime Percentage calculation, Redis may, in its sole discretion, issue a corresponding Service Credit.

4.2 RDI Exclusions. This SLA does not apply to any performance or availability issues (and such will not count toward any Downtime calculation) resulting from:

(a) Interruptions to a Workspace or Pipeline during RDI upgrades, non-Redis infrastructure failures, Source Database failures, Redis Target Database failures, Customer connectivity failures, Customer configuration issues, or collector recovery events;

(b) RDI Data in the Redis Target Database that does not reflect the most recent RDI Data from the Source Database due to (i) Customer permission changes that disrupt a Pipeline; (ii) out-of-memory conditions in the Redis Target Database; (iii) temporary disconnections between the Source Database or Redis Target Database and a Pipeline; (iv) Pipeline process recovery and catch-up after failure; or (v) lack of high availability configuration by the Customer;

(c) Changes to the Source Database IP address, endpoint, or failover topology, or Customer’s failure to implement or maintain a supported connectivity configuration for such changes;

(d) Core collector recovery within a Pipeline, to the extent caused by Source Database conditions, Redis Target Database conditions, Customer configuration, Customer connectivity, or other circumstances outside Redis’ reasonable control; or

(e) Incomplete data consistency in the Redis Target Database during the initial synchronization of a Pipeline.

5. Workload Management.

5.1 Configuration. Redis uses capacity and throughput measurement standards to assist Customer in configuring the Cloud Services based on typical usage (the "Reference Workload"). Redis will use commercially reasonable efforts to provision Customer with sufficient resources based on the Reference Workload relative to Customer's requested capacity and throughput (the "Provisioned Workload"). Customer's actual volume of data, command complexity, response sizes, and payload sizes determine the actual performance of the Cloud Services (the "Actual Workload"). If Customer's Actual Workload exceeds the Provisioned Workload, Customer may need to increase the Provisioned Workload. Redis may notify Customer if the selected plan and Provisioned Workload are insufficient for Customer's observed usage. Redis may slow or block data transfers if Customer's Actual Workload causes excessive congestion or reaches applicable limits ("Workload Restrictions"). Redis is not liable for any performance issues caused by or resulting from Workload Restrictions.

5.2 Grace Period. If Customer's Actual Workload exceeds the Provisioned Workload, Redis will notify Customer. From the date of notice, Customer has two weeks (the "Grace Period") to work with Redis to resolve the issue. During the Grace Period, Redis will support Customer's Actual Workload at no additional charge, provided the excess does not exceed 150% of the Provisioned Workload. If Customer's Actual Workload exceeds 150%, Redis may apply Workload Restrictions immediately. Customer shall respond to Redis' communications within five (5) business days; failure to do so entitles Redis to treat the Grace Period as expired. Upon expiration of the Grace Period, Redis will adjust the Provisioned Workload to reflect Customer's Actual Workload, and Customer shall pay the applicable fees from the date of adjustment. If Customer does not agree to the adjustment, Redis may apply Workload Restrictions. If Customer triggers more than two Grace Periods in any six-month period, Redis may require an immediate increase to the Provisioned Workload, apply Workload Restrictions, or suspend the SLA for the affected Commercial Subscription until the Provisioned Workload is adjusted. The SLA does not apply during any Grace Period.

6. Maintenance.

Redis may perform Maintenance at any time. Customers may select a standard weekly period for Maintenance within the Redis user interface (a "Maintenance Window"). Redis will not necessarily perform Maintenance every week. Customers who select a Maintenance Window may postpone it no more than once per month and no more than ten times per calendar year via the Redis user interface. Maintenance Windows are not available for Essentials plans. Redis will use commercially reasonable efforts to perform Maintenance within the designated Maintenance Window and to minimize the duration and impact of all Maintenance on the Cloud Services, however, Maintenance may require more time than anticipated and may complete outside the Window. If Customer has not selected a Maintenance Window, Redis may schedule Maintenance at its sole discretion. Redis will provide notice to Customer when Maintenance begins and ends. For high-impact Maintenance, Redis may provide reasonable advance notice to Customers with Annual or Pro plans. Redis will not provide advance Maintenance notice to Customers on Essentials plans. Redis may perform urgent Maintenance at any time regardless of plan type or Maintenance Windows. If Customer-side actions are required in advance of Maintenance, Customer will be given notice and reasonable time to act. If Customer fails to do so, the SLA will not apply to Downtime or performance issues caused by Customer’s failure to complete the required Customer-side actions, and Redis will not be liable for issues that could have been mitigated by following Redis' Maintenance notice instructions.

7. End of Life and Changes.

Cloud Services based on older versions may be decommissioned, become unavailable, or no longer be supported eighteen months after a newer version is made generally available, or as otherwise announced by Redis. Redis will use commercially reasonable efforts to notify affected Customers with Annual plans at least 90 days prior to the end-of-life date. Redis reserves the right to change the terms of this SLA by publishing updated terms on its website, such change to be effective as of the date of publication provided that, unless otherwise stated in the Agreement, any change will apply prospectively.