Comments on: Handling Cache Invalidation http://www.davidcramer.net/code/61/handling-cache-invalidation.html A blog about Django, JavaScript, CSS, and general web development. Mon, 08 Sep 2008 11:17:50 +0000 http://wordpress.org/?v=2.6.1 By: Ivan Jovanovic http://www.davidcramer.net/code/61/handling-cache-invalidation.html#comment-18996 Ivan Jovanovic Mon, 10 Mar 2008 12:12:47 +0000 http://www.davidcramer.net/code/61/python-and-php-memcached-client-comparisons.html#comment-18996 I'm now exploring around the problem of caching and invalidation of 1-n relations between the items. What do you think about some general mechanism of that kind of invalidation that could be provided by memcached for example. Could the relations between items and their invalidation upon change be implemented into caching mechanisms, better than into application logic. As I see this is a general problem so maybe to solve it in more general place could be more appropriate???? I’m now exploring around the problem of caching and invalidation of 1-n relations between the items.
What do you think about some general mechanism of that kind of invalidation that could be provided by memcached for example.
Could the relations between items and their invalidation upon change be implemented into caching mechanisms, better than into application logic.

As I see this is a general problem so maybe to solve it in more general place could be more appropriate????

]]> By: David http://www.davidcramer.net/code/61/handling-cache-invalidation.html#comment-15456 David Fri, 07 Dec 2007 19:40:11 +0000 http://www.davidcramer.net/code/61/python-and-php-memcached-client-comparisons.html#comment-15456 I was actually looking at .NET's caching framework -- and the SqlCacheDependency is pretty cool. We are also going to come up with a solid (hopefully) solution for dependencies. But it seems the best solution is going to be store another object in memory that gives reverse mappings to keys which are dependent on keys. You would then simply register a dependency to gain invalidation on it, but you'd automatically gain invalidation when objects were removed from the cache still using this system described above. I was actually looking at .NET’s caching framework — and the SqlCacheDependency is pretty cool.

We are also going to come up with a solid (hopefully) solution for dependencies. But it seems the best solution is going to be store another object in memory that gives reverse mappings to keys which are dependent on keys.

You would then simply register a dependency to gain invalidation on it, but you’d automatically gain invalidation when objects were removed from the cache still using this system described above.

]]> By: Al http://www.davidcramer.net/code/61/handling-cache-invalidation.html#comment-15446 Al Fri, 07 Dec 2007 06:16:18 +0000 http://www.davidcramer.net/code/61/python-and-php-memcached-client-comparisons.html#comment-15446 It might be worth reading through the documentation for the <a href="http://msdn2.microsoft.com/en-us/library/ms178604.aspx" rel="nofollow">SqlCacheDependency</a> framework for .NET to see if you can gleam any other useful information for this as well. It might be worth reading through the documentation for the SqlCacheDependency framework for .NET to see if you can gleam any other useful information for this as well.

]]>