Comments for post Redis weekly update #1 - Hashes and... many more!
adam lieus writes: This is huge info. This is really a Good site,Always I like visit this site due to its contain and useful information.
http://mailingsoftwares.blogspot.com/
mail mascot writes: I want to try this for a project, but unfortunately Redis does not work on Windows now. I want the server supported POSIX and Win32.
please visit the source : http://mailmascot.co/
adam23jones writes: Thank you very much for your tips. I will follow the steps given by you.
http://www.mailmascot.com/
Ajay writes: Just out of curiosity: why not use mkstemp() to generate the temporary filename?
Michael Stillwell writes: Oh, I didn't think it would be exposed, I was wondering more about the implementation, and any performance implications.
If I know I'm going to end up with a hash of thousands of keys, is it much more efficient to manage this myself, or is it better to leave it to Redis?
i.e. is there a difference between
HSET foo field_1 xx
HSET foo field_2 xx
...
HSET foo field_9999 xx
and
SET foo_field_1 xx
SET foo_field_2 xx
...
SET foo_field_9999 xx
if there are very many items in the "inner" hash?
(My guess is that the main differences are that "whole of hash" operations such retrieving the list of keys, and deleting the hash are easier/possible with the HSET style, but that this might come at a small performance cost?)
antirez writes: @Michael: this conversion is not exposed to the user at all. The Hash type is always perceived as a key holding an aggregate type composed of fields associated with values. The zipmap/hashtable thing is just the internal representation for this.
Michael Stillwell writes: You say that when a hash get big it is "automatically converted into a real hash table." Does this mean it becomes part of the "main" key/value store? i.e.
HSET key field value
gets converted into something like
SET key+field value
? Or does each hash have its own completely separate datastructure?
Antonio Ognio writes: So far loving Redis and this weekly update. It's very informative and definitely a win for the project. Please keep doing it!
dude writes: Oh man I quite literally laughed out loud at Nate's comment Re: Windows support. Ha! Made my morning.
litaocheng writes: great work!I love the weekly update!
Mordy writes: I love the weekly update. Please continue with it.
Nate writes: I would love to try this out for a project, but unfortunately Redis doesn't run on Windows now. I wish the server supported POSIX and Win32.
Marcus writes: I love the hash support! I was using PHP serialization, but hashes would be a lot better!
lpgauth writes: Hash support ftw! I was using JSON, this is going to simplify my life :)
home