Scalability

http://www.sdsqlug.org/presentations/May2009/MySpace_DB_Overview.pptx myspace architecture presentation, talks about how to myspace solves its database issues

http://www.slideshare.net/cstrep/myoperacom-scalability-v20 slides from my.opera.com architecture presentation. Iptables/linux lvs -> apche/mod_perl -> mysql replication, varnish, memcached

http://habrahabr.ru/blogs/hi/69457/#habracut high-load web notification system or how to do persistent connections in ajax/js. Useful for counters, friend-status updates and so on. Example is given for libevent-based daemon in perl.

http://xph.us/software/beanstalkd/ seems to a rare thing: a message queue server, running as a native code

http://highscalability.com/yahoo-s-pnuts-database-too-hot-too-cold-or-just-right writeup on yahoo PTUNS (distributed database ). Also, see the original paper if you are interested.

http://www.metabrew.com/article/anti-rdbms-a-list-of-distributed-key-value-stores/ comparative review of different DHT scalable systems

http://www.gluster.org/docs/index.php/GlusterFS fully-distributed network file system (fully = no separate meta-data servers ). Supports data redundancy (data can be stripped/mirrored across several nodes), supposed to be high performance. supports encryption and can be embedded in app server (i.e. apache won’t need to go through the kernel for FS access )

http://www.mysqlperformanceblog.com/2006/08/09/cache-performance-comparison/ gives you an idea how different caching methods perform. Read it before implementing something

http://cookiesareforclosers.com/blog/2008/06/linkedin-architecture interesing notes about LinkedIn architecture (JAVA + MYSQL + ORACLE)

http://blog.emmettshear.com/post/2008/03/03/Dont-use-Pound-for-load-balancing another evidence that if u need a loadbalancer – ngnix is the answer these days [and maybe perlball]