Hi, woud you say its best practise to save alarms/events or weight data in PI DA or use an RMDBS for that instead? Isnt DA more appropriate for analog data like temperature and so on?
I'm not sure what the best practice is, but what makes sense for you seems to be more important.
The data archive can provide long term, reliable storage for data, whether that is discrete data such as an alarm condition or weight as well as trending data like a temperature or moisture.
How long does the data need to be available will also affect the storage mechanism.
Thank for your answer!
lets say the data are from an machine weightening tablets of a rate of at least 10 per second x 10 tags. Every value needs to be saved (no exception/compression). how fast can an advise point/with its interface (lets say Opc DA) collect data and keep track of it? Then say you mix in Analyses on top of that in AF. Even tho DA may be able to handle it, Would it not be better/easier to collect it too an rbmds and link to that instead? Architecture Wise
I have little to no experience with OPC DA Advise; I have used OPC on poll extensively, so I cannot comment on advise.
In term of the data archive handling it, that is not an issue; a former co-worker worked with an interface that received 30 values per second and the data archive didn't even blink.
For the analysis engine, I don't think the issue will be the volume of data, it is very robust.
My question would be more around how is the data being processed? It is an analyses per item or is it aggregate data?
Retrieving data ...