Codecov Report

Attention: Patch coverage is 50.00000% with 2 lines in your changes are missing coverage. Please review.

Project coverage is 59.97%. Comparing base (e9e69d9) to head (9ee78f3).

@@            Coverage Diff             @@
##              dev    #2404      +/-   ##
==========================================
- Coverage   59.98%   59.97%   -0.01%     
==========================================
  Files          65       65              
  Lines        9433     9434       +1     
==========================================
  Hits         5658     5658              
- Misses       3775     3776       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

Under some conditions a 30 sec retry is too long.

Good practice is to use exponential backoff - 1 second, 2 seconds, 4 seconds, etc. up to MAX_RETRIES

So maybe replace:

await asyncio.sleep(sleep_time)

with

await asyncio.sleep(2 ** retry_counter)

and drop the sleep_time completely.

In that case it would be nice to document this exponential behaviour in the .env.example file as a comment explaining that setting FUNDING_SOURCE_MAX_RETRIES=5 will make LNbits retry after 1 second, 2 seconds, 4 seconds, 8 seconds and 16 seconds before stopping and defaulting to VoidWallet.

Good practice is to use exponential backoff - 1 second, 2 seconds, 4 seconds, etc. up to MAX_RETRIES

True, but the server does not answer any requests until it initialises the funding source.
So better to try and fail faster to VoidWallet than to be un-responsive for a longer time.

Also, this PR is a response to #2403 that removes the retry altogether.

So better to try and fail faster to VoidWallet than to be un-responsive for a longer time.

We can use

await asyncio.sleep(0.2 * (2 ** retry_counter))

so the wait will be 0.2, 0.4, 0.8, 1.6 and 3.2 seconds (total 6.2 seconds) for FUNDING_SOURCE_MAX_RETRIES=5

so the wait will be 0.2, 0.4, 0.8, 1.6 and 3.2 seconds (total 6.2 seconds) for FUNDING_SOURCE_MAX_RETRIES=5

Done, thank you!

Using 0.25 leads to less awkward numbers (sorry for suggesting 0.2 first, I only later realized that 0.25 is nicer)

👍