Disable when developing?

[Mathieson]

Is there any way to disable retries when I'm actively developing? We want Tenacity to catch everything when deployed to production, but I want to know immediately if something is broken when working and making changes. Tenacity catches and retries, though, prolonging the feedback loop.

I thought there might be an environment variable or something that would allow me to disable things, but I didn't see anything mentioned in the docs and so I thought I would ask here.

[notypecheck]

@Mathieson Depending on how you're using the library you could create all instances of Retrying/AsyncRetrying using your custom function, but monkey patching is also an option 🤔

[pySilver]

having callable parameter would be really nice addition.

@Mathieson what approach you are using?

[jd]

Fixed in #604 — added an enabled parameter to @retry. When enabled=False, the function is called directly without any retry logic:

import os

@retry(
    enabled=os.getenv("ENABLE_RETRIES", "1") != "0",
    stop=stop_after_attempt(5),
    wait=wait_fixed(1),
)
def call_api():
    ...

You can also use retry_with(enabled=False) on a per-call basis.