Make a workaround if download stucks!

[mkb79]

I think I would use publication_datetime. Seems like it would be more streamline, especially considering it seems like pre-order is new.

Thanks. I don't have any podcasts in the library but will add that option.

I've also noticed another issue, my downloads get "stuck" it seems quiet often on large download files. I am running -v debug as I was hoping it would show anything but nothing.

debug: Currently downloaded chapters: 1
debug: Currently downloaded voucher files: 1
error: Currently downloaded aycl voucher files: 1
Voucher file saved to /mnt/12tb_1/Audible RIPs/0Audible Plus RAW/Parnos_Company_The_Black_Sheep_of_Soulan_Book_1-AAX_44_128.voucher.
debug: Currently saved voucher files: 1
Parnos_Company_The_Black_Sheep_of_Soulan_Book_1-AAX_44_128.aaxc:  59%|██████████████████████████████████████████████████████████████████████████████▏                                                     | 614M/1.01G [08:00<05:33, 1.33MB/s]

This has been stuck at 614M for the past 6 hours now.

I have bad internet here, so I assume that may be the issue. Would there be a way to make the script check if the download hasn't changed in say something like 10minutes, it either retries, restarts or skips the file and throw an error so we can monitor?

Originally posted by @johnlescault in https://github.com/mkb79/audible-cli/issues/115#issuecomment-1236858862

[mkb79]

@johnlescault Which timeout you set when downloading the file?

[johnlescault]

I always set 90000, it seems to be the only way I could get it all started - maybe I should lower it?

[mkb79]

90000 means 25 hours. That’s the reason why the download stucks and don’t raise a Timeout. You can try to set a lower timeout or remove it and set a start date (e.g. 2022-09-01). This way the library list is much smaller and faster to receive.

[mkb79]

How described on the httpx docs I can set the timeout to Timeout(XX, read=600.0) where XX is the time you specified with the --timeout option. This will raise a ReadTimeout after 10 minutes. But 10 minutes is a long time.

[johnlescault]

So I did both reduced the timeout (600 I think I put) and set a start date. So far so good doing it that way.

I thought the timeout was for the fetching of the library, and not each individual downloads, so it makes sence why it did that.