holyna wrote:
Now, I have a question on how to obtain the sensor ID. As I mentioned in the beginning of the post, I'm using Fieldpoint system. I think instead of "ID", it gives "Fieldpoint IO Point Out". I assume that corresponds to the channel ID ? If so, how can I convert that "Point out" as some kind of string as the ID for the sensor? I mean, in terms of VI, how can I achieve that.
The ID can be anything you want. It does not need to be a specific "property" of the item. It can be something as simple as an increasing index. An example of how it's used is in the page that talks about JOIN statements on the w3schools site. Here you can see that each person is assigned an ID. In this case it's just an increasing index (which SQL Server can automatically create when you add new entries to a table). The Orders table links to the Persons table via the person ID, which is a column in the Orders table so you know who placed that order.
Another basic question here, if we save all data into one table, can we backup the database without interrupting the vi programming? I think it should be ok.
The issue here isn't with the VI running and trying to add records, it's with being able to back up databases. While SQL Server is running the database files are open. Most backup utilities that don't cost oodles of money will not back up open files. Some have SQL Server plug-ins, but they cost money. Here at work I have a SQL Server Agent (a tool internal to SQL Server) task that runs right before our nightly backup. The SQL Server Agent simply backs up the database to a separate directory, so the normal backup backs up that file, without having to try to deal with open files.
There's a reference talking about how to use queue function to combine different types of signals, but I can't find it right now. I'm just wondering which one is better (faster) between shared variables and queue functions.
I haven't used shared variables so I can't really say which one is better or faster.
Getting back to the DB issue, I record my data into DB in the non-deterministic loop, where I have to specify which signal goes to which column (my initial design). If we do the 2-tier design, how should I link multiple shared variables to 3 columns in term of vi programming?
You would add a row for each reading from each sensor. In other words, instead of writing data horizontally (across 200 columns) you'd add data vertically using multiple records. If we continue with the example I provided, then you'd be adding records that would require a sensor ID. You'd first need to lookup the sensor ID based on the sensor that generated the data and use that ID as the value to insert into the column of the table that holds the data.
Let's take a concrete example. Let's say you use 2 tables. One table defines sensors, called "tblSensors" or something, that looks like this:
| ID | Module | Channel | RawMeas | PhysMeas |
| 1 | cFP-AI-111 | 1 | Volt | Flowrate |
| 2 | RTD-122 | 5 | Ohm | Temp |
Now let's say you have the data table, let's called it "tblData" that looks something like this (with some dummy data):
| ID | Date | RawMeas | PhysMeas | SensorId |
| 1 | 08/16/2009 02:24:36 PM
| 0.35
| 2.3 | 1 |
| 2 | 08/16/2009 02:24:37 PM | 0.35 | 2.3 | 1 |
| 2 | 08/16/2009 02:24:38 PM | 0.36 | 2.4 | 1 |
| 2 | 08/16/2009 02:24:36 PM | 1100 | 27.1 | 2 |
| 2 | 08/16/2009 02:24:37 PM | 1200 | 27.5 | 2 |
| 2 | 08/16/2009 02:24:38 PM | 1000 | 26.8 | 2 |
The data here indicates three readings, once a second, from 2 sensors. 3 readings x 2 sensors = 6 records. When you want to add, say, a flowrate reading, then you would need to know the sensor ID that corresponds to the sensor that measures flowrate. In this case it's 1. You can get this one of 2 ways: (1) from SQL Server using a SELECT statement on the tblSensors table; (2) have a lookup table that you just keep that's synchronized with the SQL Server table.
