Server logs?
I am wondering how to inspect the application server logs.
Doing
railway logs just tails the most recent deployment's logs.
Thank you.38 Replies
Project ID:
N/AN/A
It seems that server logs are taking a long time to appear or not showing up at all. It appears that the issue is related to the ReactDevTools extension and that there is a 3-minute delay on the dashboard.
⚠️ experimental feature
Not really my use case.
you can view them on the website
or do you need to see them via the cli?
the website would be fine
but where exactly?
click on the desired deployment then click on tab for what type of logs you want to view
The deploy logs are not tailing the application server for some reason.
But the problem may be on my side, not yours.
are you wanting to view logs from an older deployment in this scenario?
I am interested only in the current deployment
then
railway logs or in the UIin the UI is fine
what shows up in the cli?
same as the UI
I am using gunicorn
as http server
okay so what seems to be the problem?
I think it's on my side actually
???
LOL
maybe gunicorn should be configured to show logs
but it's weird
I should at least get something in the deployment logs
you might be running into the python buffered logs issue but just not explaining things correctly
I am thinking about it right here with you
Let me recap
I am visiting the Deploy Logs
from either the CLI / UI
I'd expect to see some kind of output from the application server
but I don't
set
PYTHONUNBUFFERED = 1Ok, let me try.
then don't say things are fine like you've previously stated
I never said things are fine 🙂
I said that the problem might have been on my side
^^
and that I am not seeing anything in the logs because I might need to turn on logs in the app server
No, I was just answering your question about where I would have preferred seeing the logs. "The UI is fine" means "I prefer the UI, thank you".
You asked a question. 😄
mhm
I am going to try the environment variable you just suggested.
I added the env var to the Django service, automatic redeploy kicked in, but to no avail.
proof?
I wonder why it's not showing gunicorn's welcome screen
or whatever it is the one that precedes the server loop
have you setup any kind of logger or changed any logging levels?
doing it right now
that was just a question
Yes, I realized I set gunicorn to be just "warnings" in the start command.
yep that would do it
keep that env variable in there for good measure though
alright, thank you.