I use a data historian (sometimes called time series database) for collecting and persisting large (billions) of rows of  measurement data.  The data being collected is off of a manufacturing equipment and represents sensors such as temperature, pressure….  I've been wondering if I should be researching some type of BigData replacement.  The historian simply stores key=value types data, primarily made up of timestamp=value.  At 3:00, temperature was 40, at 3:01, it was 40…. Lots of repetitive data, but historians are good at compression, but cost $$.  I have to believe that commodity hardware is a lot less than year over year software maintenance.  Has anyone used any of the Apache Hadoop products in this scenario?

Thanks

I did something similar (Historian + Manufacturing Intelligence tool for
Batch Processing industries) using OpenTSDB.

On Wed, Oct 3, 2012 at 6:38 AM, Wendy Buster <[EMAIL PROTECTED]>wrote:

> I use a data historian (sometimes called time series database) for
> collecting and persisting large (billions) of rows of  measurement data.
>  The data being collected is off of a manufacturing equipment and
> represents sensors such as temperature, pressure….  I've been wondering if
> I should be researching some type of BigData replacement.  The historian
> simply stores key=value types data, primarily made up of timestamp=value.
>  At 3:00, temperature was 40, at 3:01, it was 40…. Lots of repetitive data,
> but historians are good at compression, but cost $$.  I have to believe
> that commodity hardware is a lot less than year over year software
> maintenance.  Has anyone used any of the Apache Hadoop products in this
> scenario?
>
> Thanks
>
>

Hello,

Take a look at http://opentsdb.net/overview.html it's really look like what
your are describing.

Cheers

2012/10/3 Wendy Buster <[EMAIL PROTECTED]>

> I use a data historian (sometimes called time series database) for
> collecting and persisting large (billions) of rows of  measurement data.
>  The data being collected is off of a manufacturing equipment and
> represents sensors such as temperature, pressure….  I've been wondering if
> I should be researching some type of BigData replacement.  The historian
> simply stores key=value types data, primarily made up of timestamp=value.
>  At 3:00, temperature was 40, at 3:01, it was 40…. Lots of repetitive data,
> but historians are good at compression, but cost $$.  I have to believe
> that commodity hardware is a lot less than year over year software
> maintenance.  Has anyone used any of the Apache Hadoop products in this
> scenario?
>
> Thanks
>
>
--
Damien