This morning, I saw a long test upload of mine had failed. It keeps
   failing on restart. It calls out to a tool called arkb, but fails with
   this:
   Requesting: [1]https://arweave.net:443/tx_anchor
   Error: timeout of 100000ms exceeded
   The error shows it is a simple GET request, and happens again when I
   run the process again. You can see the full error uploaded at
   [2]https://arweave.net/bxVRcuA4AIDlVqV_Z5L2tW742TQkS4pwxVnw-FZotxA .
   Meanwhile, if I do the request manually, it's working fine:
   $ time curl [3]https://arweave.net/tx_anchor
   t4cNiM_h_U_6lA0DGq5qYSb-vGGp3L7B0K0L74M6f-4LfJNZP2sHfe_nB3M3bQFF
   real    0m0.564s
   user    0m0.024s
   sys     0m0.011s
   Similarly, uploading the error message worked fine. The failure only
   happens from a specific arkb upload, and presently is happening
   reliably.
   It's frustrating that I see no immediate indication of why the error
   would occur, and would need to resort to a packet log to properly
   diagnose it. The call happens within a dependency library of arkb.

References

   1. https://arweave.net/tx_anchor
   2. https://arweave.net/bxVRcuA4AIDlVqV_Z5L2tW742TQkS4pwxVnw-FZotxA
   3. https://arweave.net/tx_anchor