Mass Data Storage on cRIO

[MrQuestion]

I'm using a cRIO system that will be saving massive amount of data and will need several terabytes of storage capacity that would be capable of having a sustainable write speed of several hundred Mb/s with a possible max burst ride speed of up to 600Mb/s.

This is raw waveform data. It is possible that this data could be significantly smaller if I take full advantage of the FPGA and run a real time compression algorithm. I have not had the time to investigate this option or run trial tests. 

I have been looking at options for the cRIO and there doesn't seem be a clear path for this solution.

The SD card on the cRIO's seem to max at 32Gb, and seem to be limited on write speed. The modules for external data on the cRIO back-plane seem to be insufficient.

I'm currently looking at using a Network Allocated Storage (NAS).

But while researching this there doesn't seem to be a clear way to mount the network drive within LabVIEW or the real time OS.

I have found in the forums where somebody was able to use a NAS that had FTP capabilities and used the FTP VI's successfully.

Using a NAS also introduces other issues such as Network jitter. But, I think this can be minimized by dedicating the second Ethernet port just for the NAS.

The issue that I'm having with this solution is that I can not find many NAS that has clear enough documentation that defines if the system supports a FTP heedlessly and/or has clear definition on sustainable write speeds.

Does anybody have experience using a NAS with a cRIO? 

I would like to hear success stories or failures? 

I would also like to know if I'm heading down the correct path from a hardware architecture perspective. 

Regards,

[BradM]

Just for complete clarity, are you talking 600 Megabits/sec (roughly 60MiB/s)?

Depending on the needs of the application (jitter-wise) and the controllers being used (does it have USB 3.0), network may be the best solution. There are some network storage options, both supported through LabVIEW directly (WebDAV) or configured for local network storage mounts (e.g. CIFS/SAMBA Windows network share access, NFS network share access, etc.).

When talking about jitter, are you concerned about network-originating jitter (i.e. you have requirements on network latency or availabilty) or application jitter? Using the different ports helps the first but not the second, while separating network communicaiton to a non-time-critical loop helps the second.