AWS Security ChangesHomeSearch

AWS wellarchitected: Restructure ADVPERF04 data management best practices and add common issues

Service: wellarchitected · 2026-11-11 · Documentation low

File: wellarchitected/latest/video-streaming-advertising-lens/advperf04.md

Summary

Rewrites the ADVPERF04 question, adds a 'Capability intent' section and a 'Common issues to watch for' list, renames best-practice links (BP02, BP04, BP05, BP06), and updates the footer link from ADVPERF03-BP03 to ADVPERF03-BP04. Content is focused on data store selection, latency/availability/cost tradeoffs, sharding/replication, observability, user profile data retention, and Clean Rooms query performance.

Security assessment

The change is a performance-pillar documentation restructure (data store selection, latency tiers, cost optimization, Clean Rooms query tuning). The only security-adjacent wording is retention policies and geographic placement for user profile data, but it is framed as a storage-cost and bid-path performance concern rather than a security control, and no vulnerability, CVE, or incident is referenced.

Evidence

+  * User profile data is managed with retention policies and geographic placement that bound storage costs while maintaining bid path performance. 

Diff

diff --git a/wellarchitected/latest/video-streaming-advertising-lens/advperf04.md b/wellarchitected/latest/video-streaming-advertising-lens/advperf04.md
index 2dd6f5e..5f044d8 100644
--- a//wellarchitected/latest/video-streaming-advertising-lens/advperf04.md
+++ b//wellarchitected/latest/video-streaming-advertising-lens/advperf04.md
@@ -6,0 +7,2 @@
+Capability intentCommon issues to watch for
+
@@ -9 +11,3 @@
-ADVPERF04: How do you choose the right data management solution for your advertising workload?   
+Advertising workloads have different data classification, availability, and access latency needs based on their subsystems. Implement a strategy to identify an appropriate AWS data store service option, along with scaling and monitoring considerations. The optimal data management solution for a particular system varies based on data type, access patterns, required throughput, frequency of access, frequency of update, and availability and durability constraints. 
+
+ADVPERF04: How do you design your data management strategy to meet the latency, availability, and cost requirements of your workload?   
@@ -10,0 +15,21 @@ ADVPERF04: How do you choose the right data management solution for your adverti
+[ADVPERF04-BP01 Choose a data management strategy that matches your availability, latency, and access requirements](./advperf04-bp01.html)  
+[ADVPERF04-BP02 Consider purpose-built databases for advertising workloads](./advperf04-bp02.html)  
+[ADVPERF04-BP03 Review your distributed database setup (sharding and replication) for performance, cost, and availability needs](./advperf04-bp03.html)  
+[ADVPERF04-BP04 Enable detailed performance and observability monitoring to tune queries and refine compute and storage](./advperf04-bp04.html)  
+[ADVPERF04-BP05 Manage high-volume user profile data](./advperf04-bp05.html)  
+[ADVPERF04-BP06 Optimize privacy-enhanced collaboration queries for performance and service limits](./advperf04-bp06.html)  
+  
+## Capability intent
+
+  * Every datastore in your advertising stack is selected based on a clear understanding of the latency tier, availability requirements, and access patterns of the data it holds. 
+
+  * Real-time bid data, audience segmentation data, and historical reporting data each land in a service optimized for their specific characteristics. 
+
+  * Your distributed database setup (sharding and replication) is designed for your performance, cost, and availability needs, not applied uniformly regardless of data category. 
+
+  * You have detailed performance and observability monitoring configured to tune queries and refine compute and storage allocation. 
+
+  * User profile data is managed with retention policies and geographic placement that bound storage costs while maintaining bid path performance. 
+
+  * [AWS Clean Rooms](https://aws.amazon.com/clean-rooms/) query performance is optimized through data partitioning, aggregation, and filtering before query execution. 
+
@@ -12 +36,0 @@ ADVPERF04: How do you choose the right data management solution for your adverti
-Advertising workloads have different data classification, availability, and access latency needs based on their subsystems. Implement a strategy to identify an appropriate AWS data store service option, along with scaling and monitoring considerations. 
@@ -14 +37,0 @@ Advertising workloads have different data classification, availability, and acce
-###### Best practices
@@ -16 +39 @@ Advertising workloads have different data classification, availability, and acce
-  * [ADVPERF04-BP01 Choose a data management strategy that matches your availability, latency, and access requirements](./advperf04-bp01.html)
+## Common issues to watch for
@@ -18 +41 @@ Advertising workloads have different data classification, availability, and acce
-  * [ADVPERF04-BP02 Consider purpose-built and streaming databases](./advperf04-bp02.html)
+  * Using a single database technology across all advertising data categories regardless of their latency profile, forcing low-latency real-time bidding data and high-latency historical reporting data to compete for the same storage resources. 
@@ -20 +43 @@ Advertising workloads have different data classification, availability, and acce
-  * [ADVPERF04-BP03 Review your distributed database setup (sharding and replication) for performance, cost, and availability needs ](./advperf04-bp03.html)
+  * Storing all historical user profile data indefinitely in DynamoDB, paying provisioned or on-demand pricing for data that is rarely queried during bid evaluation. 
@@ -22 +45 @@ Advertising workloads have different data classification, availability, and acce
-  * [ADVPERF04-BP04 Enable detailed performance and observability monitoring to help tune queries and refine compute and storage](./advperf04-bp04.html)
+  * Replicating user profile data across multiple regions in real time, incurring high cross-region transfer costs for data that is almost never accessed from more than one Region simultaneously. 
@@ -24 +47 @@ Advertising workloads have different data classification, availability, and acce
-  * [ADVPERF04-BP05 Manage high volume user profile data](./advperf04-bp05.html)
+  * Deploying bidder nodes and database nodes in different Availability Zones without Availability Zone-aware routing, paying cross-AZ data transfer charges on every database query from the bid path. 
@@ -26 +49 @@ Advertising workloads have different data classification, availability, and acce
-  * [ADVPERF04-BP06 Consider AWS Clean Rooms collaboration](./advperf04-bp06.html)
+  * Running Clean Rooms queries against large unpartitioned datasets without aggregations or filters, hitting the 5 GB result size limit and generating long query processing times. 
@@ -37 +60 @@ To use the Amazon Web Services Documentation, Javascript must be enabled. Please
-ADVPERF03-BP03 Use a cloud file system to store shared data between applications
+ADVPERF03-BP04 Use single-zone high-performance object storage for latency-sensitive ad serving and model artifact retrieval