Indexing on a project hosted via SSHFS makes pycharm unusable...disable indexing?
Running Mac OS X Lion, I have a project that is mounted on a SSHFS drive. The amount of time it takes for indexing makes pycharm pretty much unusable for this project. Is there any way to fix this?
Please sign in to leave a comment.
Such configuration is not supported. Your project files should be located on the fast local drive. You can use built-in Deployment feature to synchronize with the remote server.
Yeah, that constraint is hardly unique to PyCharm. I'd tried using sshfs for WingIDE and even Emacs, with poor results. Any IDE that's I/O-intensive is going to have this problem, and you can't do all the magic stuff without I/O.
PyCharm's Deployment feature is decent if you turn on automatic uploads. Otherwise you either have to copy everything over, or keep track of what you changed to copy.
Something I have done successfully, however, is run an IDE on a remote server and tunnel X11. That worked reasonably well over a 100Mb LAN. I don't know if you're a *nix user, but if you are, X11 is a surprisingly decent option for remote development, assuming you don't mind installing GUI software on your devel server.
Currently, the source (a Subversion checkout) is located on a remote (LAN) Linux VM. My desktop computer is a Windows XP machine. This is where I run PyCharm. I have the source directory (on the Linux machine) shared via Samba and mapped on my Windows machine. I "simply" treated it as local and had PyCharm import the project. It seemed to work great, and I've been using it this way for a few months, BUT..I suffer with performance issues all the time and near constantly spinning tasks. I've suspected this was related to accessing the source via a mount over the network.
I'm reading about "Server Configuration" and "Deployment" and "Remote Interpreters", but I'm not clear on how these concepts apply to my situation exactly.
It appears I could copy the source to a local drive on my Windows machine. Then I could setup my Linux machine as a "Server" in Pycharm–a machine I "deploy" to. However, what would happen to my Subversion integration? Subversion is configured on my Linux machine. So rather than copy the source over from the Linux machine, would I need to install Subversion on my Windows machine and checkout the source direct to my local Windows drive so I can import this "new" project into PyCharm?
I don't want to ask too many questions in one post...but...assuming I'm correct with the above...am I correct to think the "deployment server configuration" would ensure my local source files will be kept in sync on my Linux machine (the server). How seamless is this syncing? Like a lot of folks, when working to debug javascript, for example, will make a change, save, and immediately hit refresh in the browser to see the affect of the change. In my case, the browser will be running against the web server on my Linux machine.
Getting my source local seems to have solved all the problems I was having. Mclagett, I'd be open to discussing how your journey is going on this, too.
However, I'm having some line-ending issues related to working with code on Windows and deploying to a Linux machine. I describe the situation here:
http://forum.jetbrains.com/thread/PyCharm-1141