What does ReadFull: unexpected EOF mean when forgetting --prune

When running a forget --prune on a B2 hosted repository, i.e.

forget --prune --keep-within 48h --keep-daily 14 --keep-weekly 8 --keep-monthly 6 --keep-yearly unlimited

I saw this in the output:

repacking packs
Load(<data/ab1860ecea>, 6515913, 0) returned error, retrying after 716.576042ms: ReadFull: unexpected EOF
Load(<data/ab1860ecea>, 6515913, 0) operation successful after 1 retries
Load(<data/7b8f11d335>, 17312091, 0) returned error, retrying after 561.803955ms: ReadFull: unexpected EOF
Load(<data/7b8f11d335>, 17312091, 0) operation successful after 1 retries
[1:05] 100.00%  128 / 128 packs repacked

What is this? B2 throttling? I have a pretty rock solid business class internet connection.

Apparently the TCP connection to B2 was interrupted twice for some reason. The cause of the interrupted network connections can be located anywhere between your host and the B2 servers. The most likely cause is that a few hickups occurred somewhere. As long as those errors don’t occur regularly you can probably ignore them.