implement the 3x amplification limit #2536
Conversation
marten-seemann
force-pushed
the
amplification-protection
branch
2 times, most recently
from
3bc2e98
to
5248e4f
Compare
| func (p *packetPacker) packCoalescedPacket(buffer *packetBuffer, maxPacketSize protocol.ByteCount) (*coalescedPacket, error) { | ||
| maxPacketSize = utils.MinByteCount(maxPacketSize, p.maxPacketSize) | ||
| if p.perspective == protocol.PerspectiveClient { | ||
| maxPacketSize = protocol.MinInitialPacketSize |
There was a problem hiding this comment.
So as client we're always ignoring the parameter? Should we make that clearer, e.g. maxPacketSizeForServer?
There was a problem hiding this comment.
Only for coalesced packet sent by the client. This doesn't apply for normal packets.
| @@ -1427,7 +1428,7 @@ func (s *session) sendPacket() (bool, error) { | |||
|
|
|||
| if !s.handshakeConfirmed { | |||
| now := time.Now() | |||
| packet, err := s.packer.PackCoalescedPacket(protocol.MaxByteCount) | |||
| packet, err := s.packer.PackCoalescedPacket(s.sentPacketHandler.AmplificationWindow()) | |||
There was a problem hiding this comment.
Is there a risk for busy-looping here, if we exit the send loop when it's time to send an ACK? Or is this the same as if we're congestion limited, and we'll deal with that gracefully?
There was a problem hiding this comment.
This should be handled the same way as being congestion limited.
marten-seemann
force-pushed
the
amplification-protection
branch
from
5248e4f
to
fae5068
Compare
This will trigger the amplification protection.
marten-seemann
force-pushed
the
amplification-protection
branch
from
fae5068
to
e33f7d0
Compare
Codecov Report
@@ Coverage Diff @@
## master #2536 +/- ##
==========================================
+ Coverage 86.19% 86.22% +0.03%
==========================================
Files 122 122
Lines 9583 9603 +20
==========================================
+ Hits 8260 8280 +20
Misses 985 985
Partials 338 338
Continue to review full report at Codecov.
|
Fixes #1862.