Apply suggestion from @v-albemi

Author: v-albemi

articles/mysql/flexible-server/concept-performance-best-practices.md

@@ -82,7 +82,7 @@ While partitioning has its benefits, it also has some limitations, such as the l
 
 Most applications primarily read from the database, with only a small percentage of the interactions involving writes. The number of active connections on the primary database we calculated for the connection pools likely includes read traffic. Offloading as many queries to read replicas as possible and conserving access to the primary writable instance increases the amount of overall database activity performed by the application servers without increasing the load on the primary database. If you aren't already accessing read replicas at least for longer running queries such as reports, you should consider immediately moving reporting or analytics to read replicas.
 
-Wider scale use of read replicas might require more careful consideration, as replicas lag slightly behind the primary because of the asynchronous nature of replication. Find as many areas of the application as possible that can be served with reads from the replicas with minor code changes. You should also apply this method at higher levels concerning caching - serve more of the read only or slowly changing content from a dedicated caching tier such as [Azure Managed Redis]( ../../azure/redis/overview.md).
+Wider scale use of read replicas might require more careful consideration, as replicas lag slightly behind the primary because of the asynchronous nature of replication. Find as many areas of the application as possible that can be served with reads from the replicas with minor code changes. You should also apply this method at higher levels concerning caching - serve more of the read only or slowly changing content from a dedicated caching tier such as [Azure Managed Redis](/azure/redis/overview).
 
 ## Write scaling and sharding