dockerless issue
I am using runpod serverless. To reduce deployments, I store the server executable files in a network volume, as described in the link https://blog.runpod.io/runpod-serverless-no-docker-stress/.
I have been using this method for several months, but recently, even when I modify the rp_handler.py file in the network volume, the changes do not seem to be reflected immediately and appear to be cached. As a result, I am currently unable to use it properly.
Are there any recent changes regarding this issue?
RunPod Blog
Deploying Python Machine Learning models on RunPod, without any doc...
What if I told you, you can now deploy pure python machine learning models with zero-stress on RunPod! Excuse that this is a bit of a hacky workflow at the moment. We'll be providing better abstractions in the future!
Prerequisites and Notes
* The tutorial only works for containers installed purely
7 Replies
What if you redeploy ?
Like setting it to 0 workers then 3 again
Might need to reach support for this like on the website
you have to make sure to do a new release, otherwise old stuff can stay running
changing an env variable is enough to trigger a new release
So you mean that rp_handler.py in network handler is cached? any update in runpod logic?
Yes probably it's all in a running container with its own data so it's like cached
You might have to redeploy it like that
Is it intentional that the network volume is cached differently than in the past? I don't know why there was a need to introduce a cache to the network volume. Very uncomfortable.
Because, every time I change a file in my network volume, I have to change the environment variable and initialize the cache. This causes more delay because it re-initializes docker internally at the same time.
Not sure if its cached really, or its running all the time in background and not updated on file update
that should of been same before, its not cache but works cache as they run, you need to do a release to shuffle workers incase 1 stays running for a while