In putobject.cpp, the normal flow where boost http is reading bytes from the network (the body) and writing to iRODS or a local part file, the read was updated to be asynchronous instead of synchronous. This was necessary because using both async and sync causes conflicts on the socket.
This same change needs to be made in that file when we are manually parsing the chunk header.
In putobject.cpp, the normal flow where boost http is reading bytes from the network (the body) and writing to iRODS or a local part file, the read was updated to be asynchronous instead of synchronous. This was necessary because using both async and sync causes conflicts on the socket.
This same change needs to be made in that file when we are manually parsing the chunk header.