ERROR: connect: connection refused (ECONNREFUSED) #33
Comments
|
Will you do I think I'm relying on some un-tagged changes to Blink.jl |
|
It's strange... It works now without changing Blink. |
|
Strange indeed. I'll close now, but feel free to reopen/post if it resurfaces |
|
I got the same error and then when I used Any idea what is causing this? |
|
I've seen this come up a few times and haven't been able to figure out just what causes it. Were you running this julia session remotely, perhaps via SSH? |
|
FYI I frequently get this too. I think it's some sort of race condition during window setup for Blink. When I re-run the command it usually works. |
|
Yes, Spencer, I was running julia via Putty on a remote server. Might that have something to do with the problem? I've run the command a few times consecutively and it always gives out the same error message. |
|
@i-dash yes, that would be a problem. What is happening is that Blink.jl starts up a web server on the remote machine, but the remote machine blocks communication to it for security reasons. This is a good thing. If you ssh somewhere, you don't really want that computer to be opened up on any port. We still have not figured out a way to work remotely with Blink. |
|
I also got this error, managed to solve forking Blink and changing default interval of function try_connect, solving the race condition! Original: Modified: Any solution to keep original code of Blink allowing to update it and override the function?! |
|
@AntonioLoureiro that's interesting. Instead of changing interval can you change attempts to something like 300 and let me know if the issue goes away? |
|
Changed attempts to 300 and it also works! And this solution is better i think!! Do you know any elegant manner to override this arg not forking Blink Module? |
|
@MikeInnes I'm thinking of just changing the default in blink -- are you opposed to that? |
|
Please do. In general we probably need a better approach to making this connection. |
|
OK great, thanks. @AntonioLoureiro that has been merged into Blink and a new release has been tagged on METADATA. Whenever the metadata PR is merged, you should be able to go back to the released blink and have this change! |
|
@MikeInnes @sglyon I am having the same connection refused error and have the new Blink.port = 5382 and I an running Any thoughts or suggestions? Thank you! |
|
I still see this issue on Ubuntu 16.04. I use Julia in a local terminal, versioninfo: Playing around with the default arguments of Has there been any progress on this? |
|
Hi @krcools sorry to hear that. I haven't heard of anyone having this issue when using a terminal with a connected display on a local machine. Can you post more of the output of the Julia session when the issue came up? |
|
Hi and thanks for the quick response. Yes, given that I'm the only one seeing this and that extending the time-out settings does not fix it, it is probably something different. The stack trace is very similar to the above though. Here is a complete session, from Julia start-up to the error being thrown: |
|
Hmm. This is strange indeed. Can you tell me what happens when you try that same code from within a Julia session started with If you don't have the |
|
The same thing happens. |
|
I should probably take my issue to Blink. Going through there example gives me the same result... Thanks for your help. |
|
Running Blink.AtomShell.install()usually solves the problem for me |
|
A complete reinstall of Julia and the .julia package directory fixed the problem for me. |
|
Thanks for the report and sorry for the inconvenience! |
|
I want to do simulations in Julia with the "Plots" package and when compiling I get this error, someone who has solved it can help me: connect: connection refused (ECONNREFUSED) Stacktrace: When compiling the PlotlyJS package, this error appears: Plotly javascript loaded. To load again call init_notebook(true) WARNING: Method definition (::Type{PlotlyJS.SyncPlot{TD} where TD<:PlotlyJS.AbstractPlotlyDisplay})(PlotlyJS.Plot{TT} where TT<:PlotlyJS.AbstractTrace) in module PlotlyJS at /home/toodivz/.julia/v0.6/PlotlyJS/src/display.jl:136 overwritten at /home/toodivz/.julia/v0.6/PlotlyJS/src/displays/ijulia.jl:187. MethodError: no method matching PlotlyJS.GenericTrace(::String; showlegend=true, mode="lines", xaxis="x", colorbar=Dict{Symbol,Any}(Pair{Symbol,Any}(:title, "")), line=Dict{Symbol,Any}(Pair{Symbol,Any}(:color, "rgba(0, 154, 250, 1.000)"),Pair{Symbol,Any}(:shape, "linear"),Pair{Symbol,Any}(:dash, "solid"),Pair{Symbol,Any}(:width, 1)), y=[0.0998334, 0.109778, 0.119712, 0.129634, 0.139543, 0.149438, 0.159318, 0.169182, 0.17903, 0.188859 … 0.00681464, 0.0168139, 0.0268115, 0.0368064, 0.0467976, 0.0567841, 0.066765, 0.0767392, 0.0867057, 0.0966635], name="y1", yaxis="y", x=[0.0, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09 … 6.19, 6.2, 6.21, 6.22, 6.23, 6.24, 6.25, 6.26, 6.27, 6.28]) Stacktrace: |
|
Hey I'm getting this old error again. On a fresh install (new laptop) of julia 0.6.3 on ubuntu 18.04, upon
I ran Any ideas? |
|
CIBot seems to hit this consistently so it would be great if somebody knew how to avoid this error. |
|
Does CIBot run with a "fake display" (e.g. inside xvfb-run)? |
|
Yes. It uses |
|
Does it perhaps spawn a subprocess or connect to a remote process that doesn't have a fake display? |
|
The current code is basically @schedule run(`Xvfb :99`)
sleep(10)
com = Cmd(`$julia -e "$test_cmd"`; env=("DISPLAY"=>":99",))
run(com)and I also tried with just |
I'm getting the following error in two different computers:
This one is:
The text was updated successfully, but these errors were encountered: