You're right about the memory consumption for large files being due to Capistrano's interface. That could certainly be improved. But the wild CPU usage and slow upload speed is due to Net::SFTP (and to some extent, Net::SSH). Whatever you can come up with would be appreciated, certainly!
- Jamis On Jun 7, 2007, at 8:50 AM, Mat Schaffer wrote: > On Jun 7, 2007, at 10:39 AM, Jamis Buck wrote: >> That would be Net::SFTP's fault, and I'd certainly welcome any >> patches that made Net::SFTP more optimal. A word of caution, >> though...the Net::SFTP code is...messy. It's pretty hard to follow, >> and I say that as the author of the library. I'm pretty embarrassed >> by both Net::SSH and Net::SFTP. > > Huh... that's not what I was expecting. I had assumed that the > nastiness was due to the fact that put took the full file contents as > an argument rather than streaming the file. I'll poke around a bit > more and send in patches if I have anything worth-while. > > I just didn't want to start work that was doomed due to some kind of > best-practice that I didn't know about. > > Thanks again, and didn't realize you wrote Net::SSH and Net::SFTP, so > thanks for that too. The interface kicks way more butt than any > other SSH library I've tried. Implementation can be fixed, api is > trickier. So I think you've done a fine job so far! > > -Mat > > > --~--~---------~--~----~------------~-------~--~----~ To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/capistrano -~----------~----~----~----~------~----~------~--~---
