Test Valkey Helm 1.0 Before GA!

Ruby developers now have a generally available GLIDE client built on the same shared Rust core used by GLIDE’s Java, Python, Node.js, and Go clients. The release includes proactive reconnection, automatic cluster topology handling, multiplexed connections, and native OpenTelemetry tracing. Developers migrating from redis-rb get familiar API conventions, although the new client is not a drop-in replacement.

Valkey Admin 1.2 adds a cluster topology utilization view that brings memory usage, CPU, operations per second, hit ratios, and connected clients into one interface. Nodes exceeding 90% memory or 85% CPU utilization are flagged, making potential bottlenecks easier to identify. The release also introduces safer network defaults and tighter authentication for metrics collectors.

LLM inference repeatedly processes context that may already have been computed. In an experiment using Valkey to store and reuse KV cache data, the same two GPU nodes handled 25% more requests while streaming responses twice as fast. The results demonstrate how externalizing cached attention state can improve GPU utilization and inference throughput without adding GPU capacity.

FakeValkey lets Python developers test Valkey commands entirely in memory, eliminating the need to start a server or container for every unit test. Built into fakeredis, it supports common operations including keys, expiration, transactions, and scripting. The guide demonstrates how to integrate it with pytest, test migrations, and identify workloads that still require real-server integration tests.

Valkey monitoring starts with collecting the right metrics. This tutorial explains how to expose Valkey performance data to Prometheus, visualize it through Grafana, and choose an exporter based on deployment requirements. It provides a practical starting point for tracking performance, understanding resource consumption, and improving observability across production deployments.

Inspecting production data requires more care than running commands against a development instance. This guide explains how to use SCAN and INFO alongside read-only ACL permissions to examine keyspaces while reducing operational risk. It also explores how LibreDB Studio approaches safe data inspection and access control for running Valkey instances.