You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@hbase.apache.org by Amit Sela <am...@infolinks.com> on 2012/05/17 10:02:18 UTC

Unique ID per URL

Hi all,

One of our HBase tables holds URL as a row key.

I read that it is recommended to hold the URL key as: reversed domain +
 URL ID (using unique id per url.)

I understand the part of reversed domain but could anyone elaborate about
unique id per url, maybe give an example ?

Thanks.

Re: Unique ID per URL

Posted by sagar naik <sn...@attributor.com>.

We use md5(url)
That gives us a good distribution

-Sagar

On Thu, May 17, 2012 at 1:02 AM, Amit Sela <am...@infolinks.com> wrote:

> Hi all,
>
> One of our HBase tables holds URL as a row key.
>
> I read that it is recommended to hold the URL key as: reversed domain +
>  URL ID (using unique id per url.)
>
> I understand the part of reversed domain but could anyone elaborate about
> unique id per url, maybe give an example ?
>
> Thanks.
>