Redis gets a cache:evict command on Commerce Cloud

Adobe Commerce Cloud adds a cache:evict command, handing merchants direct control over Redis cache memory instead of leaving it to automatic policies.

Redis gets a cache:evict command on Commerce Cloud

Adobe has shipped performance work for Adobe Commerce Cloud aimed at Redis caching. The headline is a new cache:evict command, which gives you control over what sits in cache memory and when it goes.

Why the cache matters

Redis caching underpins Adobe Commerce Cloud performance. It cuts database queries and expensive computation, so pages come back faster. The trade-off is memory — caches grow, and on a store with a large catalogue and a lot of business logic they grow quickly.

Until now, managing that meant leaning on Redis's built-in eviction policies to remove old data automatically. That works, but it makes its decisions without knowing anything about how an eCommerce platform actually uses its cache.

What cache:evict does

The cache:evict command moves the decision to you. Instead of waiting for an automatic policy to fire, you choose what to remove and when — keeping the data that earns its place in memory and clearing the rest. Two things follow: lower memory consumption, and better cache hit rates, because what's left is the data you wanted cached in the first place.

Running it

The command works with existing Adobe Commerce Cloud deployments. You can run it in a maintenance window or wire it into an automated process, and the syntax is plain enough that it doesn't need a specialist. Scheduling eviction for off-peak hours, or after a specific application event, keeps cache memory in check without touching customer-facing traffic.

What merchants are seeing

Deployments that have optimised their cache report faster page loads, better checkout performance and steadier response times under load. The effect is largest on stores with big catalogues or heavy customisation — the ones generating the most cache overhead to begin with. Clear out what isn't earning its keep and there's more headroom for serving real requests.

Make it a routine, not a reaction

Get the most from it by treating cache management as a regular job: scheduled eviction of non-critical entries, monitoring of cache memory usage, and adjustments based on what you observe rather than what you assumed. Track consumption over time — growth trends tell you where to optimise before performance degrades in front of a customer.

What this means for you

If you're on Adobe Commerce Cloud, cache strategy is now something you decide rather than something Redis decides for you. Worth putting on the agenda at your next platform review.