Revision history [back]
In eppy
, idf.run()
will run on any machine that has eppy installed and E+ installed.
Service like Heroku (Platform as a service) generally will NOT allow you to install E+
Considering this limitation, my solution has been to:
Run it two machines
- Run the web server on Heroku (or something similar).
- Install and run E+ on a separate machine on AWS or linode.com (call it the E+Machine)
- The Webserver sends the input files over the internet to the E+Machine to be simulated
- Once the simulation is complete, the E+Machine sends the results back to the web server that does something with the results
Take a look at zeppy. It may help you with steps 3 and 4 above
Running it on a single machine
You can run everthing on a single machine on ASW or linode.com - both webserver and simulation. You lose the ease of deployment that "Platform as a service" Heroku gives you. But it makes everything else simpler, as you don't have to deal with the communication betwen two machines.
In eppy
, idf.run()
will run on any machine that has eppy installed and E+ installed.
Service like Heroku (Platform as a service) generally will NOT allow you to install E+
Considering this limitation, my solution has been to:
Run it two machines
- Run the web server on Heroku (or something similar).
- Install and run E+ on a separate machine on AWS or linode.com (call it the E+Machine)
- The Webserver sends the input files over the internet to the E+Machine to be simulated
- Once the simulation is complete, the E+Machine sends the results back to the web server that does something with the results
Take a look at zeppy. It may help you with steps 3 and 4 above
Running it on a single machine
You can run everthing on a single machine on ASW or linode.com - both webserver and simulation. You lose the ease of deployment that "Platform as a service" Heroku gives you. But it makes everything else simpler, as you don't have to deal with the communication betwen two machines.