Stata with AWS S3 or Google Buckets

[Daniel Feenberg]

There is interest here in cloud computing, and I wonder if anyone in this forum is doing much with Stata in the cloud. The stumbling block I see is that S3 objects and Google Buckets aren't readable or writable by Stata. Is that right? So each time a user needs a file they have to copy it from the object store to a local temporary partition, and remember to copy back any new files? The alternative of using "persistent storage" seems to negate the potential savings. I don't see that any other packages are better at this, but my experience is limited. Am I missing something?

[Mauricio Caceres]

Would using a FUSE mount-point help? It's not local so it wouldn't take any space on your system. You can use and store files by simply using "/path/to/mount-point/bucket-name/file.dta".