Tuning Memcached Timeouts for a Cloud Environment

As applications increasingly migrate to cloud environments, the use of memcached is adapting to these new settings, exemplified by nearly 300 applications utilizing NorthScale's memcached service on Heroku’s cloud platform. Unlike traditional LAN environments where server failures were more common than network issues, cloud networking is less controlled, often resulting in packet loss and high latency due to shared resources and multiple node hops. This shift necessitates adjustments in handling timeouts for memcached operations; aggressive timeouts may lead to failed queries if packet loss occurs, as TCP’s initial retransmit timer is three seconds. It is suggested to loosen server failure limits and adjust timeout settings to accommodate network-related timeouts rather than server issues. Particularly for applications not originally designed for cloud environments, tuning timeouts to manage at least a single dropped packet is recommended, with specific parameters for connect and retry timeouts suggested for the Fauna Ruby client recommended by NorthScale.

No tracked trend matches for this post yet.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.