want to change algorithm used in OPP for token and key comparison

[Sagar Agrawal]

Hi,

I am using OrderPreservingPartitioner, and my keys are integers which are
stored as strings, I want to manually assign token values equal to my key
values such that data is equally distributed.
So for this to work, I want to convert the token and key strings to integers
before doing compareTo ( I suppose OPP does a simple compareTo between key
and token value).
So I think I will have to change the code and re compile it.
Would it be a tedious task to do, I don't want to miss out anything.
Any suggestion would be highly appreciated?

Thanks

nodetool move gives exception

[Sagar Agrawal]

When I did nodetool move , nodetool streams for that node shows
mode:bootstrapping
but modetool move gave following exception
http://gist.github.com/471503

but the mode is still bootstrapping...

what does that mean?

Thanks

Re: manual InitialToken assignemnt

[Sagar Agrawal]

got it, thanks

On Fri, Jul 9, 2010 at 6:21 PM, Per Olesen  wrote:

> Are you using OrderPreservingPartitioner or RandomPartitioner?
>
> Cause if you are using RandomPartitioner, a hash is calculated from "a" and
> that hash is used to determine where the data for "a" key goes, not "a".
>
>
> On Jul 9, 2010, at 2:16 PM, Sagar Agrawal wrote:
>
> > I have a 2 node cluster
> > node1 - 5
> > node2 - 9
> >
> > If I insert a row with key="a", which node should it go and why?
> >
> > It is going to node1, but I think it should go to node2, since token
> value of node is closer to "a" (using java string compareTo method)
> >
> > someone please clarify
> >
> > Thanks
>
> Med venlig hilsen/Best regards
>
> Per Olesen, Developer
> Trifork A/S
>
> Spotorno Alle 4, DK-2630 Taastrup
> Mobile: +45 23389581 - Mail: p...@trifork.com<mailto:p...@trifork.com>
>
>

manual InitialToken assignemnt

[Sagar Agrawal]

I have a 2 node cluster
node1 - 5
node2 - 9

If I insert a row with key="a", which node should it go and why?

It is going to node1, but I think it should go to node2, since token value
of node is closer to "a" (using java string compareTo method)

someone please clarify

Thanks

Re: total disk space used on a node for a CF is too large than expected

[Sagar Agrawal]

what does WriteCount signify actually, it should also include writes which
are replicas right? It is total no of writes on that node for that CFtill
now, right ?

On Fri, Jul 9, 2010 at 2:39 PM, Sagar Agrawal  wrote:

>  row size is 10 KB and write count on a node for a CF is 1054451,
> so ideally the total disk space used on that node by that CF should be
> around 10 GB
> but it's showing  23 GB
> what else might be taking up so much space?
>
> Thanks
>

total disk space used on a node for a CF is too large than expected

[Sagar Agrawal]

 row size is 10 KB and write count on a node for a CF is 1054451,
so ideally the total disk space used on that node by that CF should be
around 10 GB
but it's showing  23 GB
what else might be taking up so much space?

Thanks

error in I/O Console updater

[Sagar Agrawal]

I am populating data into my cassandra db using hector. There are about 9
million rows, and I have given enough heap space n all but after writing 2-3
lacs rows, it shows an error,

This is the error log. http://gist.github.com/462345

Thanks

Re: searching keys of the form substring*

[Sagar Agrawal]

Thanks Vineet for replying, but I am not able to understand how can we use
variable substitution in it.



On Mon, May 31, 2010 at 4:42 PM, vd  wrote:

> Hi Sagar
>
> You can use variable substitution.
> ___
> Vineet Daniel
> ___
>
> Let your email find you
>
>
>
> On Mon, May 31, 2010 at 3:44 PM, Sagar Agrawal  wrote:
>
>> Hi folks,
>>
>> I want to  fetch all those records from my column family such that the key
>> starts with a specified string...
>>
>> e.g.  Suppose I have a CF keyed on full names(first name + last name) of
>> persons...
>> now I want to fetch all those records whose first name is 'John'
>>
>> Right now, I am using OPP and KeyRange in the following way:
>>
>>  KeyRange keyRange = new KeyRange();
>> keyRange.setStart_key("John");
>> keyRange.setEnd_key("Joho");
>>
>> but this is sort of hard coding can anyone suggest a better way to
>> achieve this?
>>
>> I would be really grateful... thank you.
>>
>>
>>
>

searching keys of the form substring*

[Sagar Agrawal]

Hi folks,

I want to  fetch all those records from my column family such that the key
starts with a specified string...

e.g.  Suppose I have a CF keyed on full names(first name + last name) of
persons...
now I want to fetch all those records whose first name is 'John'

Right now, I am using OPP and KeyRange in the following way:

 KeyRange keyRange = new KeyRange();
keyRange.setStart_key("John");
keyRange.setEnd_key("Joho");

but this is sort of hard coding can anyone suggest a better way to
achieve this?

I would be really grateful... thank you.