RT File Size issues

[abales]

We are running PXI 8101 RT on LabView 2011 August release. 

Occasionally we find the free space is gone and our data dir is corrupted.

We are storing 50k/Sec*8channels for 8 hours in TDMS format.  This is probably about 80+GBytes.  We were doing it in 1 TDMS file.  We have been doing 1,2,4G byte files now.  However sometimes we find the free space is gone and the files are corrupted or destroyed, requiring a reformat and reinstall of the RT.

We are finding that we can do 1-2Gigabyte files forever.  The software checks for free space, deletes some old data, and starts a new file.  It checks to make sure it has 100G free.  This works.  We are putting the data into a subdir under www so its easy to find for ftp'ing.

I suspect we can only do up to, but not including, 4Gigabytes for one TDMS file.  Is this true?  Or do we have some other issue?  TIA

[SnowMule]

Limitation's probably in the disk's file system, not the file format.  FAT, NTFS can really only play nice with <2GB files; much bigger than that and you're asking for trouble.  Not saying it can't be done, but you've got to keep the disk formatting in mind with a system like that.

[abales]

your right.  RT is FAT32 and its max'd at 4G per file. 

So I was wrecking the File alloc table (FAT) and blowing up the system.  Thought about using a external USB drive that was NTFS but RT won't recognize it per NI's website.  Looks like we are just going to do 20 minute chunks of TDMS files (3.6G) and that probably helps with the FTP as well.  80G files won't FTP well anyways.

I wonder if anyone will really look at all of that data anyways 😉

[SnowMule]

Is the test system on a network?  Could you stream the data to another machine with a listener that can write to a large disk array?

But yeah, breaking the test data down into smaller chunks is an easy way around it.