RIO1900 memory gets full after 3 days the application is running

[alois007]

I wrote a RIO1900 application to collect meteorological data like temperature and air-pressure and save it to a USB-stick.

The application is working, but after about three days the physical memory (readout with NI MAX, starting at about 90 Mbyte) of the RIO1900 is decreasing to zero, than the "Status" LED is blinking four times in red colour.

Who knows this problem and who can solve it?

[GerdW]

Hi Alois,

---

@alois007 wrote:

I wrote a RIO1900 application …

Who knows this problem and who can solve it?

---

It's you who write this application - and it should be you who can solve the problem!

Generic answer: don't collect ever-growing array data inside your VI(s)…

[Bob_Schor]

I assume the "RIO1900" is what NI (and I) calls a "myRIO".  This is intended to be the "Target" side of a LabVIEW Real-Time Project, with a PC running Windows and an appropriate version of LabVIEW (such as 2019 32-bit or 2021 32-bit), with a compatible version of the myRIO Software Toolkit, and the myRIO running NI Linux Real-Time OS, with communication between the two machines utilizing TCP/IP, typically over a USB connection.

The myRIO has a limited amount of nonvolatile memory, 512 MB, used for the OS and local storage, and an additional 256 MB of volatile memory.  When using this as a data-acquisition device, with the intent of saving the data, you should probably consider collecting the data on the myRIO (which has fast and flexible A/D and digital I/O and streaming it to the Host (with typically GB or TB of storage).  I use Network Streams for this, others have used TCP or UDP protocols, often utilizing a Producer/Consumer pattern to transfer the data with minimal interference to the Real-Time processing taking place in the myRIO.

Bob Schor