Closed saeedAdelpour closed 7 months ago
Hi @asherf I wanted to kindly ask if you could spare a moment to review my recent pull request? Thank you so much
hi, thanks for the PR. I think this is not a good change in the context of this package. it complicates the code. the app can run this type of code instead of calling this package. I think this type of code in the package (setting up an endpoint) is mostly for reference. ideally, the app should define its own endpoint function and implement it in a way that fits the app specific reqs.
Hi,
Thank you guys for your very powerful library, You are GREAT!
When we run gunicorn command to serve django app, the command is
gunicorn APP.wsgi:application --name APP --workers 4 --bind=0.0.0.0:8000 --log-level=INFO --log-file=-
then 4 process runs this part of code, one of them allocates metrics on the
PROMETHEUS_METRICS_EXPORT_PORT
and if one of them allocates, this is enough for the rest, because all other processes write their metrics inPROMETHEUS_MULTIPROC_DIR
i had this workaround in my django to solve this issue
this workaround dosen't work on celery processes, we need to run these
export PROMETHEUS_MULTIPROC_DIR=PATH/TO/
celery ....
when we set this env variable on OS layer, we can use this code in our backend
I still checking python client to solve the issue,