Memcached Roundup: Memcached 1.4 Released, Gear6's WebCache

Posted by Mirko Stocker on Jul 27, 2009

Sections

Architecture & Design,

Development

Topics

Ruby ,

Memcached ,

Dynamic Languages ,

Distributed Cache ,

Caching ,

Clustering & Caching ,

Languages ,

Scalability ,

Programming ,

Infrastructure ,

Performance & Scalability

Share |

RELATED Research

Memcached, a caching system originally developed for LiveJournal.com has been adopted by many sites to lower the load on their database servers.

The recently released version 1.4 introduced some new features, like the new binary protocol. Why a new protocol? Memcached developer Dustin Sallings says:

The most obvious thing for protocol implementors is that it’s now really easy to parse the protocol. After reading a fixed-size header, a low-level packet processor can figure out where to dispatch the input and split it into all of its major components (key, value, opaque, cas, extras, etc...).

That's great for the (small) number of developers who write servers and clients, but what about random people out there who just want to use memcached? Semantic enhancements in the new protocol allow us to build some really cool stuff.

One example of this is Trond Norbye's replication feature, "where you store only a few of the items on multiple servers".

Because applications talk to Memcached via a standardized and open protocol, implementations of the Memcached back-end can be interchanged. A commercial alternative is WebCache from Gear6:

Gear6 Web Cache is an advanced, protocol-compliant Memcached solution that enables web site developers and operators to scale web services and applications.

InfoQ had a chat with Bill Takacs, Gear6's director of product management to learn more about WebCache, how it works and what benefits it provides:

Gear6 provides a number of enhancements to standard memcached.  These include:

1. Memory utilization: Removal of the 1MB object size limit,  finer grained block based memory allocation, and a cost based eviction algorithm.
2. Density: We use a combination of DRAM and Flash memory to lower the cost of the cache and increase the density of our solution. Currently our largest cache is 384GB per 1U.
3. High Availability: We deploy our solution with two 1U units in a cluster environment. The cluster enables two modes:
   1. Continuous service availability: The cluster architecture enables fail-over capabilities. This ensures that cache services are not interrupted when failures occur.
   2. Continuous data availability: The cluster replicates data within the cluster. This replication ensures that all cache data is always available in an alternate location, and continues to be served to users without interruption or delay. Spikes in database and application load are avoided.
   3. In addition the Gear6 Web Cache requires no client-side code modification and our cluster architecture enables disruption-free software upgrades.
4. Reporting and Management: Gear6 Web Cache is fully instrumented and equipped with intuitive interfaces that let you see what’s happening at multiple levels within your Memcached tier. We’ve made enhancements that automatically and continuously scan both DRAM and flash memory for failures or failure indicators. Users can drill-down on any level of their cache tier and get reports on hot keys, clients and instances.

So how do you know whether WebCache will solve your performance problems?

Our typical process includes on-site evaluations. Prospects are encouraged to evaluate the solution in their environment

See the Memcache tag for more information on Memcached from InfoQ.