Making plumber API available over internet

Question

I am fairly new to the plumber package in R. I got a working API to run locally on my machine and I can access it from a live JS application on the web with the code:

r <- plumb("my_api_code.r")
r$run(host = "0.0.0.0", port = 8000)

I have read that it is not a good idea to host the r code on say my personal laptop, but I just wanted to be able to run the above code, open an R server session, and then access this over the internet from a different machine for demonstration purposes.

Is there a way to access this api over the internet based on the IP address of the machine running the R session? I wasn't able to find documentation for the r$run r$host r$serve functions that are available.

Create a VM or container on a hosting service (Azure, AWS, Google, etc) and run your app there. — Hong Ooi, Mar 17 '19 at 20:39
Could you expand on your answer? Do you mean host the R code on one of these services? I know the ultimate answer is to host the R code on R Studio Connect, Digital Ocean, Docker.... but I'm just trying to show a demonstration of what plumber can do without setting up/paying for a hosting service. Also, we have a hosting service for our website but I'm not sure how to install and run an instance of R on our server. — Mxblsdl, Mar 17 '19 at 21:26

score 2 · Accepted Answer · edited Feb 24 '22 at 02:13

You can try this this plot/API available over the internet

plumbr.R

#' @get /plotly
#' @serializer htmlwidget
plotlygraph <- function(){
 library(plumber)
 library(plotly)
Animals <- c("giraffes", "orangutans", "monkeys")
SF_Zoo <- c(20, 14, 23)
LA_Zoo <- c(12, 18, 29)
data <- data.frame(Animals, SF_Zoo, LA_Zoo)
 p <- plot_ly(data, x = ~Animals, y = ~SF_Zoo, type = 'bar', name = 'SF Zoo') %>%
   add_trace(y = ~LA_Zoo, name = 'LA Zoo') %>%
   layout(yaxis = list(title = 'Count'), barmode = 'group')
 return(p)
}

Run this above code using the following command on R Console

r <- plumber::plumb("plumbr.R")
r$run(host="0.0.0.0", port=8000)

After running this The Swagger window will open There is button "Get/Plotly" click on this then click on "try out" click on "Execute" wait for few seconds

Step 1:

https://i.stack.imgur.com/mEcg4.png

Step 2:

https://i.stack.imgur.com/EapIw.png

Step 3:

https://i.stack.imgur.com/h74C6.png

Step 4:

[https://i.stack.imgur.com/9CkGc.png][4]

then it shows HTML code above this, it gives a Request URL like this

http://127.0.0.1:8000/plotly

copy that link paste in the browser you will see the plot

Then using terminal/command prompt get your IP...

Assume your IP is "192.168.10.04"

copy it and replace it with this like:-

http://192.168.10.04:8000/plotly

Then copy that link and paste it to your IFRAME of another web app

You will see your R Plot on a new framework (web app)...

I think this should help you

Thanks for the detailed response. You have a typo in that the script.r file should be named plumb.r or vise versa. — Mxblsdl, May 17 '19 at 04:29
@Mxblsdl I am sorry for a small mistake. I changed the file name as **plumbr.R** Now it looks better. Follow Me on Github: [link]https://github.com/maheshKulkarniX — Mahesh Kulkarni, May 18 '19 at 08:30
This will just give you access on the same network not on the internet. — Abhishek Agnihotri, Mar 24 '21 at 12:52

score 1 · Answer 2 · answered Mar 21 '19 at 19:35

It should work beautifully given you

know the ip of your laptop
the ip is accessible to the others (in your team), i.e. you are on the same network, or your computer has a globally valid ip
no firewall blocks port 8000
CORS is enabled by something like

#' @filter cors
cors <- function(req, res) {
   res$setHeader("Access-Control-Allow-Origin", "*")
   plumber::forward()
}

Obviously, this is not the answer for a production environment but works well for just showcasing/testing.

I think this is the correct answer, although I have not gotten it to work yet. The operation times out. I think it might be an issue with the firewall. — Mxblsdl, Mar 27 '19 at 17:40

Making plumber API available over internet

2 Answers2

Linked