mirror of
https://github.com/reconurge/flowsint.git
synced 2026-03-12 01:44:42 -05:00
Non-Local Version #1
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @matt-gorman on GitHub (Oct 24, 2025).
Is there a current configuration or plans to configure this in a way that would allow this to be set up in a non-local manner, so that multiple team members could look at the project? I had made some quick changes and was routing through NGINX to try and set it up that way, but because it uses FastAPI for authentication from the main app pages, requests fail due to mixed content being blocked in most browsers.
@dextmorgn commented on GitHub (Oct 24, 2025):
Hey Matt,
It should be suitable for a non-local manner, but for now only a local config is provided. The app is designed so that multiple users can be part of multiple investigations and thus share sketches and analysis with other "local" contributors.
You might want to look at
originsarray here https://github.com/reconurge/flowsint/blob/main/flowsint-api/app/main.py and update it accordingly:This list indicates the allowed origins to make requests to the api.
Making a non-local deployment config is definetly on the whishlist, so let me know if you make any progress !
@PhineusPogo commented on GitHub (Oct 27, 2025):
Hello,
I have tried adding IPs on origins but it's not working. It seems that it also miss the 5001 opening port on docker-compose.
Even if i add 5001 port on docker-compose, i have error "failed to fetch" when i try to create an account or login.
Do you have any ideas about what needs to be changed ?
@dextmorgn commented on GitHub (Oct 27, 2025):
Hey @PhineusPogo,
What exactly are you trying to achieve ? Did you manage to make it run locally first ?
@PhineusPogo commented on GitHub (Oct 27, 2025):
I try to install it on local server and exposed it to other device on same network.
Yes locally it works well when i join it on localhost.
@dextmorgn commented on GitHub (Oct 27, 2025):
Hosting should need a dedicated documentation, but to give you an idea, you would probably need to update a few components.
if you try to reach the server with curl from your other device:
If your server is reachable, you should get a 401, otherwise you would get a connection error.
Let's check that your server responds first, as the docker-compose should normally expose port 5001.
@PhineusPogo commented on GitHub (Oct 27, 2025):
Please find the result below (both commands are launched on hosting server) :
192.168.X.X as IP of hosting server
curl http://192.168.X.X:5001/docs
curl: (7) Failed to connect to 192.168.X.X port 5001 after 0 ms: Couldn't connect to server
curl http://127.0.0.1:5001/docs
RESULT OK
@dextmorgn commented on GitHub (Oct 28, 2025):
If you ran it with
make prodormake dev, it should work fine as it's using--host 0.0.0.0, making the service accessible on all network interfaces including the local IP.So if you still cannot access it, we could investigate more. Could you check that it's not a docker issue and make sure the container is correctly exposing the port:
If you see
0.0.0.0:5001->5001/tcpthe port is exposed and is likely another issue that we could start working on : network, device, etc.@PhineusPogo commented on GitHub (Oct 28, 2025):
Hi,
The port is well exposed with command make prod
@dextmorgn commented on GitHub (Oct 28, 2025):
What type of OS are you exposing the service from ? Is it a linux machine, MacOS ?
@PhineusPogo commented on GitHub (Oct 28, 2025):
It's a Debian 12
@dextmorgn commented on GitHub (Oct 28, 2025):
Great, the port is exposed, so it's maybe a network issue.
run:
You should see something like
LISTEN 0 128 0.0.0.0:5001orLISTEN 0 128 [::]:5001.Then check your firewall:
If it's active, try allowing the port:
@PhineusPogo commented on GitHub (Oct 28, 2025):
Ufw is inactive.
@dextmorgn commented on GitHub (Oct 29, 2025):
@PhineusPogo,
Your request is not left behind, I'll try to reproduce your issue on my side. It's likely a network issue, unrelated to the services themselves, at least for now.
Once we know services can be reached, we'll focus on updating the allowed origins part.
If you make any progress on this, please keep in touch here ! 🙂
@rboxem commented on GitHub (Oct 31, 2025):
Im having the same issue. After changing the Vite config to allow connections from non local host I can reach the webpage from other clients on my network, but when I try to sign in it says: "Failed to fetch".
Localhost works fine, but normally I run all my docker apps on a headless server.
@dextmorgn commented on GitHub (Oct 31, 2025):
Hey @rboxem,
Have you modified allow list in
flowsint/flowsint-api/app/main.pytoo ?Try changing the
originsarray to accept requests from any location:Then re-launch the container and see if the problem persists.
@rboxem commented on GitHub (Oct 31, 2025):
@EliottElek Yeah I already added my 'debian' host with ports to the main.py. I now commented it out and changed it to allow all like you propose, but still 'Failed to fetch'. Let me know if you need any logfile or something in which you might expect more information. btw (just FYI) after any changes i'm deleting the containers and 'make prod' again.
@dextmorgn commented on GitHub (Oct 31, 2025):
@rboxem I'll try to reproduce it on my side and come back to you, I'll let you know ! Very likely, your problem and @PhineusPogo's are related.
@dextmorgn commented on GitHub (Oct 31, 2025):
@rboxem if you
curl http://<your-server-ip>:5001from your one of your clients, what do you get ?@rboxem commented on GitHub (Oct 31, 2025):
{"detail":"Not Found"}%Thats from 2 different clients, same result
@dextmorgn commented on GitHub (Oct 31, 2025):
Great, your service is indeed reachable, so we can exclude the network issue.
I manage to make it work on my network, with a couple of changes that I just pushed :
docker-composefiles +vite.config.tsapi values to use only theVITE_API_URLfro the.env.originsto["*"]in flowsint-api/app/main.py(you can leave like that or adapt to your setup for safety reasons)Also :
Good thing you mentionned it, as deleting the containers it not enough to see changes, running
make prodwill only recreate and re-run the containers if the images are already created. But in your case the images are unchanged, so you need to also re-create the images. I would suggest running :stop the containers and delete images:
Then pull latest fixes:
Then remove the
.envfiles:Update
env.example:Finally, run
make prodagain, and let me know !@rboxem commented on GitHub (Oct 31, 2025):
Its working! Only minor detail now is that it doesnt resolve based on hostname, but it does with the IP adress in the browser.
@dextmorgn commented on GitHub (Oct 31, 2025):
Great to hear !
Do you have a local DNS ? Do you use it behind a proxy ?
@rboxem commented on GitHub (Oct 31, 2025):
No and no, but its time to set one up I guess ;).
! Before closing this issue. I'm testing differences for the local host and the client.. the client can not add any items. The localhost can.
Error on client:
@dextmorgn commented on GitHub (Oct 31, 2025):
Yep, I just found out too and just fixed it a couple minutes ago (issue https://github.com/reconurge/flowsint/issues/24).
just like before, rebase and rerun and you'll be good to go !
Thanks for pointing it out.
@rboxem commented on GitHub (Oct 31, 2025):
Working for me as well. Thanks!
@PhineusPogo commented on GitHub (Oct 31, 2025):
Working for me too ! Thanks @EliottElek