cRIO-SQL-instruments or cRio-SQL plus cRio-instruments

[Steffen01]

hello,

if lucky I get to do some instruments in Labview to display data from the testrig.

Is it better, write wfrom the cRio to SQL on a server and then extract the real time data from there, or display the data directly from the cRio. Multiple users can have multiple instruments, will this be too much load on the cRio? Just guessing, could be up to 24 sensors on it. But one instrument may just want to display one or two sensors. And the timesteps could be different. And if I change the name of the database, it would not matter, as the inputs on the cRio dont change.

[RandallCrandall'sCandleHandle]

Hi Steffen01,

This depends on a few different things. How much processing would you like to do on the data before displaying it? If you are planning on doing a lot, it might be better to offload that from the cRIO. Also, would you like to be able to display different front panels for data you select? It might be best to send the data over the network and access it from different computers. It sounds like network shared variables might work for your application. These allow you to publish data on the network and access it from your host computer. Take a look at those here:

http://www.ni.com/white-paper/4679/en/

Cheers,

Ryan C.

[Steffen01]

I think it will be a SQL database with indexed timestamps.Then the cRio just pushes the data onto the server and the server has to deal with it.

Processing depends how linear the sensors are. Really just lots of pressure sensors, flow meters and power meters are read and then there will be a few front panels say for power, flow rates, pressure drops and so on. Not really heavy processing, unless the cRio has to make up for the nom-linearities of the sensors. I guess there will be just one reading per seconds, its heat exchangers, things are not moving that fast.

I could then save bits to excel and the whole data is in the SQL file anyway.

[RandallCrandall'sCandleHandle]

Hey Steffen01,

In terms of processing that amount of data, I don't think the cRIO should have any problems. The challenge that I see would be in hosting multiple front panels on the cRIO. This forum post goes into a bit of detail about your different options for this:

https://forums.ni.com/t5/LabVIEW/crio-to-host-multiple-remote-front-panels/td-p/3055794

I also think you should be aware of the following thread when deciding how to set up your database communication:

https://forums.ni.com/t5/NI-Linux-Real-Time-Discussions/Linux-RT-accessing-SQL-Database/m-p/3604795

It goes over some of the methods you can use as well as common issues. It will probably be better to stream the data to a host PC before you attempt to write to the database since there is more built in control for database operations under LabVIEW in Windows.

Cheers,

Ryan C.

[Steffen01]

Yes, I guess, that is whats going to happen, the receiving server will feed the SQL database, this will be done in a Labview program.

I found this, indexing timestamps

http://www.sqlteam.com/forums/topic.asp?TOPIC_ID=200191

so then the LV frontends don't put too much load on the SQL database reading the latest entry. So the frontends should actualise (is this a word) kind of real time.

[Steffen01]

does each frontend need SQL server to access the database? I guess MYSQL would be the better option then