cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ramzi Rabah <rra...@playdom.com>
Subject Re: Cassandra data distribution and configuration settings
Date Tue, 17 Nov 2009 17:58:31 GMT
InitialToken is a String for OrderPreservingPartitioner and numeric
for RandomParitioner.
It is the token taken by the node on the ring in the hash space.
You want to set it evenly distributed for each node, across your hashspace.

On Tue, Nov 17, 2009 at 9:41 AM, Jon Graham <sjcloud22@gmail.com> wrote:
> Hello TR,
>
> Thanks for your quick reply :-)
>
>  My 128 characted key values are distributed evenly across the "00 ..." to
> "ff ..." data range.
>
> I did not set the InitialToken value.  Is this a starting key value for each
> node?
>
> Is the InitialToken specified a string?
>
> Thanks,
> Jon
>
> On Tue, Nov 17, 2009 at 9:19 AM, TuxRacer69 <tuxracer69@gmail.com> wrote:
>>
>> Jon,
>>
>> what is your key distribution?
>> did you set
>>
>> <InitialToken/>
>>
>> on each node according to this distribution?
>>
>> TR
>>
>>
>> Jon Graham wrote:
>>>
>>> Hello Everyone,
>>>  I am running a 10 node Cassandra cluster using a recent cassandra trunk.
>>>  After inserting a million data records using  keys of 128 character
>>> lower case hex digits and
>>> 1024 data values, I noticed that most all of the is accumulated on only 3
>>> nodes.
>>>  I am using the OrderPreservingPartitioner setting in the
>>> storage-conf.xml file.
>>>  What do I need to take to get a balanced data distribution?
>>>  Is there a way to limit the size of an individual database file?
>>>  Is there additonal Cassandra configuration documentation outside of the
>>> storage-conf.xml file?
>>>  Thanks for your help,
>>> Jon
>>>
>>
>
>

Mime
View raw message