Ticket #7 (closed enhancement: fixed)

Opened 5 years ago

Last modified 3 years ago

Fragment downstream packets depending on firewall limits

Reported by: yarrick Owned by: yarrick
Priority: major Milestone: 0.5.0 "iPassed"
Version: 0.3.4 Keywords:
Cc:

Description

If the network enforces a 512 byte limit, then that should be detected during startup. A big mtu should still be usable, and any packets too large should be sent in parts just like it works upstream right now. This is far better than lowering the mtu to like 220 to be sure to get under the limit, since the average packet size will be closer to the limit, thus increasing throughput.

Change History

comment:1 Changed 5 years ago by yarrick

  • Milestone set to 0.5.0

comment:2 Changed 5 years ago by yarrick

  • Milestone changed from 0.5.0 "Tea Online" to 0.6.0

comment:3 Changed 4 years ago by yarrick

  • Owner set to yarrick
  • Status changed from new to assigned

Created packet.[ch], moved all packet handling out from iodine.c to it, rev [605]

comment:4 Changed 4 years ago by yarrick

Converted outpacket in iodined.c to use packet_functions, created send_chunk fcn. [616]

comment:5 Changed 4 years ago by yarrick

In [618] I added support for 1 byte going from client to server in every ping/data message telling the server the lowest byte of the latest query id that contained data payload.

comment:6 Changed 4 years ago by yarrick

  • Owner yarrick deleted
  • Status changed from assigned to new

All patches above have been reverted.

comment:7 Changed 3 years ago by yarrick

  • Milestone changed from 0.5.0 to 0.4.3

Milestone 0.5.0 deleted

comment:8 Changed 3 years ago by yarrick

  • Owner set to yarrick
  • Status changed from new to assigned

I have some new ideas how to fix this, Im working on it now..

comment:9 Changed 3 years ago by yarrick

Added upstream sequence numbers in [726] and [727]

comment:10 Changed 3 years ago by yarrick

Added send_chunk in [739]

comment:11 Changed 3 years ago by yarrick

Fragmentation working in [742], currently with a static fragsize of max 1200b.

comment:12 Changed 3 years ago by yarrick

After [744], [745] client can now set his max downstream fragment size for the session with -m. Not setting it will result in autodetection being done.

comment:13 Changed 3 years ago by yarrick

  • Status changed from assigned to closed
  • Resolution set to fixed

Fixed autoprobing of max downstream fragsize in [752] and [753]. Finally.. it's done!

Note: See TracTickets for help on using tickets.