cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kant Kodali <k...@peernova.com>
Subject Re: which one of the following choices is more efficient?
Date Wed, 26 Oct 2016 08:40:26 GMT
@Benjamin Roth I think you kinda of answered my question. my use case is
really that I need to store a SkipList like structure into one or multiple
tables. I could map the entire skip list like structure into one table or I
could map a list at every level into another table and the number of levels
are not necessarily logN in my case because I may divide the number of
nodes at each level by 1/4 or 1/8.

On Wed, Oct 26, 2016 at 1:24 AM, Kant Kodali <kant@peernova.com> wrote:

> @Ali hmm..I didn't mean to say I store the same data across two tables and
> neither tables are dependent for me to do the writes twice. you can imagine
> tables are independent and the data is unique across tables but I could
> store data from multiple tables into one table. otherwords I am just
> restating my first question.
>
> On Wed, Oct 26, 2016 at 1:19 AM, Ali Akhtar <ali.rac200@gmail.com> wrote:
>
>> You would need to do each write twice and data will take up twice the
>> space as its duplicated in two places.
>>
>> On Wed, Oct 26, 2016 at 1:17 PM, Kant Kodali <kant@peernova.com> wrote:
>>
>>> I guess the question can be rephrased into "What is the overhead of
>>> creating and maintaining an additional table?"
>>>
>>> On Wed, Oct 26, 2016 at 1:12 AM, Ali Akhtar <ali.rac200@gmail.com>
>>> wrote:
>>>
>>>> Depends on the use case. No one right answer.
>>>>
>>>> On Wed, Oct 26, 2016 at 1:03 PM, Kant Kodali <kant@peernova.com> wrote:
>>>>
>>>>> If one were given a choice of fitting all the data into one table vs
>>>>> fitting the data into two tables while say (keeping all the runtime and
>>>>> space complexity for CRUD operations the same in either case)  which
one
>>>>> would you choose and why?
>>>>>
>>>>>
>>>>
>>>
>>
>

Mime
View raw message