ReadMeFeed / Community knowledge
Real questions. Useful conversations. Find the people who know your stack.
Databases & messaging · #memcached
Avoid cache stampedes when scaling Memcached. This guide compares modulo sharding vs. consistent hashing (Ketama/HRW) to minimize key redistribution during cluster changes.
Learn how Memcached’s slab allocator can cause evictions even when total memory appears free, and how to diagnose and mitigate the issue.
Memcached timeouts almost always trace to four causes: a dead daemon, exhausted memory, a blocked port, or a saturated connection limit. Here's the five-minute diagnostic sequence and the fix for each.
A diagnostic guide to memcached LRU eviction problems: how to confirm the condition with stats counters, isolate capacity, TTL, item-size, and slab causes, and apply the right fix.
Learn how Memcached’s compare‑and‑swap (CAS) operation enables safe concurrent updates, with a worked telnet example, limits, and practical verification steps.
Stop premature Memcached evictions by tuning slab allocation. Learn how to identify internal fragmentation and align slab classes with your data payloads.
Memcached's slab allocator trades memory utilization for predictable latency. Here's how slab classes work, why eviction is per-class, and how to measure your actual waste with stats slabs.
Learn how to diagnose and fix cache miss spikes and thundering herd problems in McRouter-managed Memcached clusters using consistent hashing and request collapsing.
Memcached evicts items despite free RAM because its slab allocator rounds values to fixed chunk sizes. Learn to read stats slabs, spot per-class starvation, and use slab_automove or manual reassignment to reclaim wasted pages.
linux
facebook