New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Channel layers stop consuming messages from Redis #145
Comments
I had this issue too a couple of weeks ago too. The only workaround I found was to use python 3.6. |
I've just tried to reproduce this locally and it works perfectly, though I am also on Python 3.6. Could you try on 3.6? 3.5 has a few problems with async to the point where we may have to stop supporting it. |
Confirm that upgrading python to 3.6 solve the problem. Not sure, is it correct to close this issue, because python 3.5 declared as supported version. |
I think I might just say 3.6 is required then. |
I just tried the test example with Python 3.5.4, and it seems to work. |
@d-s-e If you could pin this down to a specific change in 3.5.x that fixes the behaviour, and we can demonstrate that, then yeah, maybe we could lower the requirement. Having said that, it's swimming against the tide at this point, so it'd have to be convincing and I'm not sure how long we could maintain it going forward... |
Hm, I didn't want to dig that deep ;-) |
Sure... but that's not quite enough to go on to say that we DO support the version, only to open up a whole raft of support requests and issues. It's just not worth it, we need to be sure. |
Yes, you're right. But isn't just jumping direclty to 3.6 because "it works" just the same? |
Hi @d-s-e. It's more about capacity at this point. Unless someone can do the leg work we just can't revisit a previous decision here, especially given that Python 3.5 is now on the wane. (Django's just dropped it going forward, a recent survey puts it at just 11% usage, and so on.) Sorry. |
I fully understand the capacity problem, I can see it with my own time. I just thought, this could help others, having the the same issue. But I'm fine with that, thanks. |
I am facing exactly this issue, though on
Those who have somehow solved their issue, which minor version of |
It works well on Python 3.6.4. Just in case you have issues with 3.6.4 too,
of the other packages you listed, my versions are:
channels==2.1.6
channels-redis==2.3.2
Django==2.1.4
requests==2.21.0
I don't use mysql.
…On Wed, Feb 27, 2019 at 12:56 PM darkfishy ***@***.***> wrote:
I am facing exactly this issue, though on Python 3.6.1, with following
installed packages (including their dependencies):
channels==2.1.7
channels-redis==2.3.3
django==2.1.7
mysqlclient==1.4.1
requests==2.21.0
Those who have somehow solved their issue, which minor version of Python
3.6.y did you use?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#145 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AJ4OZ1tDUuNffDEXz6R7VKSTOoNedlnTks5vRnJrgaJpZM4ZRY_Q>
.
|
@nichoski That was pretty quick response :-) I'll test with 3.6.4 and report back as soon as I can. Thanks. |
I tested with |
Any suggestions to work that around with python 3.5.3 (current Python version in Debain 9)? |
I'm still experiencing this w/ both python related bits from pip freeze:
|
Hi!
I've experience a problem similar to #1191
To reproduce this problem I've build a sandbox project based on chat example from official tutorial, you can see it here: https://github.com/bashlakov/django-channels-bug-demo
Steps to reproduce the problem:
Open http://localhost:8000/chat and create a new chat.
Open this chat in another tab. You send message in one tab and see it in this and in second tab.
Create a new different chat in new tab and send message to it.
Come back to first chat tab and continue to send messages. After sending several messages the chat brakes down and you can't send any new massages.
At this point you can browse you Redis and see, that messages stored in specific.* channels queues but not consumed by channel layers.
There is a video with this steps: https://youtu.be/DTFE-Up1De0
I've tried this example on Mac and Linux, using latest version of packages and python 3.5.2.
I've also tried to start the project with ./manage.py runserver, daphne and even uvicorn with no luck.
I found in-memory channel layers backend in sources and check again using this implementation instead of channels_redis. I can't reproduce bug with in-memory backend, so I think the problem is related to channels_redis.
Can you please try to reproduce this problem in your environment and confirm issue?
The text was updated successfully, but these errors were encountered: