Data Historian Speed

[Shawn Ebersole]

I was just curious on which type of data historians are considered fast at retrieving historical data.  For example, if I wanted to trend a tag at a one second interval to review an event that took place a year ago and compare that to the same tag in real-time, what historian does this the fastest?

[keznz]

Hi Shawn, I think any historian should manage that with ease. The fact that data is a year old should be irrelevant with any good historian. When it is stored at the time of recording, it is indexed, which allows data that's 1 year, 2 years, or 10 years old be retrieved in roughly the same amount of time as data collected yesterday. Any differences between good historians should be negligible.

[Scadadoctor]

Shawn, We typically use Microsoft SQL Server as a short term & long term historian. You also can use it to store & retrieve data at a very high speed & then compare with real time data values also. We have seen responses like sub second to get hundreds of tags from old archived files that may have years old data.  Also we have stored thousands of tags worth of data streaming into MS SQL Server at fairly high acceptable performance speed. The traditional non Microsoft historians have a lot of overhead & are very heavy on use of system resources, so they require a lot powerful machines. Performance is same. MS SQL server is almost free, runs on a very low end machines & MS provides awesome analytical tools with it. We can share examples if you like. Good luck & cheers

Sonam