Details
-
Improvement
-
Status: Resolved
-
Major
-
Resolution: Abandoned
-
3.0.0-DEAD
-
None
Description
Currently, when we process the Handshake, we send as many buffer as we have NEED_WRAP events. If two NEED_WRAP events are consecutive, we could certainly use the same buffer, as each buffer is 16Ko big. That would save 2 byteBuffer allocation plus 32 Ko per handshake.