cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hiller, Dean" <>
Subject Re: Advice on correct storage configuration
Date Mon, 01 Oct 2012 14:12:14 GMT
What is really going to matter is what is the applications trying to read?
 That is really the critical piece of context.  Without knowing what the
application needs to read, it is very hard to design.

One example from a previous post that was a great questions wasŠ
1. I need to get the last 100 requests no matter which user
2. I need to get the last 100 requests of a specific user

This gives anyone on the list an idea of how the model should look AND
also is why on the web you will find many many many references that noSQL
is designed from the queries.  (that said, I have seen one project where
denormalization caused them to write 1 meg of data on one request so there
is a balanceŠ.and boy was that slow on the writes).


On 10/1/12 7:27 AM, "Lewis John Mcgibbney" <>

>I wish to confirm whether the current mapping (storage) configuration
>I have is suited to store data commonly extracted field data from Web
>My mapping can be seen here [0] which basically specifies three column
>families e.g. parse (p), fetch (f) and super columns (sc) within the
>webpage keyspace.
>Each column family subsequently includes several fields which for
>clarity include comments. Current CF configuration is as follows:
>- fetch CF includes 11 columns
>- parse CF including 4
>- super column CF including 7
>I am trying to ascertain why the 7 super column fields are currently
>configured to be super columns as oppose to standard columns!
>I therefore wonder if someone can please clarify if such a
>configuration is suited to storing data of this nature.
>Thank you in advance. if this is too vague an explanation the please
>say so and I will be happy to expand on any aspect in an attempt to
>fully understand the data model and the configuration.
>Thank you

View raw message