readstata13 icon indicating copy to clipboard operation
readstata13 copied to clipboard

Inconsistent read times

Open vkhodygo opened this issue 2 years ago • 3 comments

I've been trying to improve reading times of some moderately large .dta files (about 5GB each), however, those times sometimes differ by an order of magnitude. The actual value is about 6 minutes (I use tictoc for this), however, sometimes it get stuck and I have to wait for about 2 hours to get it done (with barely any disk activity).

Please, let me know what extra information you need in addition to this.

vkhodygo avatar Jun 13 '22 13:06 vkhodygo

Hi @vkhodygo , well there isn't really something on our end I can think of. Maybe some other process is blocking your memory or the CPU cycle or the disk. We simply read a file from the first to the last byte in sequencial order.

JanMarvin avatar Jun 13 '22 14:06 JanMarvin

This is extremely odd: I've rebooted the machine for the sake of consistency, Same files, same order, same code, read speed is somewhere around 1-2MB/s which is somewhat unacceptable.

vkhodygo avatar Jun 13 '22 15:06 vkhodygo

Could you please provide the output of sessionInfo(). Though I still don't think that it's us causing the slowdowns. Do you have the file on your local desktop? Is there some Antivirus running and interfering? It might as well be your station failing.

JanMarvin avatar Jun 13 '22 15:06 JanMarvin