jrandom authored 20 years ago and zzz committed 20 years ago

* add new back-off logic to reduce payload resends during transient
  lag - only let one packet be resent at a time, even if the window size
  allows it (and the packet timers request it).  this should make
  congestion less painful, and reduce the overall number of messages
  resent (as the SACKs for the one packet actively resent should clarify
  what made it through)

jrandom authored 20 years ago and zzz committed 20 years ago

    * Clean up the propogation of i2psocket options so that various streaming
      libs can honor them more precisely

jrandom authored 20 years ago and zzz committed 20 years ago

    * Minor logging update
(toss net.i2p.router.JobQueueRunner=WARN in /configlogging.jsp to see wtf is hanging your router)

jrandom authored 20 years ago and zzz committed 20 years ago

    * Fix a long standing leak in I2PTunnel (hanging on to i2psocket objects)
    * Fix a leak injected into the SimpleTimer
    * Fix a race condition in the tunnel message handling

jrandom authored 20 years ago and zzz committed 20 years ago

    * Added throttles on how many I2PTunnel client connections we open at once
    * Replaced some buffered streams in I2PTunnel with unbuffered streams, as
      the streaming library used should take care of any buffering.
    * Added a cache for some objects used in I2PTunnel, especially useful when
      there are many short lived connections.
    * Trimmed the SimpleTimer's processing a bit

jrandom authored 20 years ago and zzz committed 20 years ago

* properly implement the connection timeout
* make sure we clear the outbound packets on close
* don't b0rk on repeated close() calls

jrandom authored 20 years ago and zzz committed 20 years ago

    * Added throttles on how many I2PTunnel client connections we open at once
    * Replaced some buffered streams in I2PTunnel with unbuffered streams, as
      the streaming library used should take care of any buffering.
    * Added a cache for some objects used in I2PTunnel, especially useful when
      there are many short lived connections.
    * Trimmed the SimpleTimer's processing a bit

jrandom authored 20 years ago and zzz committed 20 years ago

allow throttles on the number of streams participated in, as well as how a timeout + queue for overflow

jrandom authored 20 years ago and zzz committed 20 years ago

tweak around with the connection delay and timeout activity to test delayed vs. immediate syn

jrandom authored 20 years ago and zzz committed 20 years ago

* allow loading the connection options from the env (or another Properties specified)

jrandom authored 20 years ago and zzz committed 20 years ago

    * Allow loading the (mini)streaming connection options from the
      environment.
    * More defensive programming in the DSA implementation.

jrandom authored 20 years ago and zzz committed 20 years ago

* ack properly on duplicates
* set the message input stream buffer large enough to fit the max window (duh)

jrandom authored 20 years ago and zzz committed 20 years ago

* make sure we ack duplicate messages received (if we aren't already doing so)
* implement a choke on the local buffer, in case we receive data faster than its
  removed from the i2psocket's MessageInputStream (handle via packet drop and
  explicit congestion notification)

jrandom authored 20 years ago and zzz committed 20 years ago

* dont increment # unacked packets artificially
* dont try to push data after closing
* cleanup the packet serialization
* logging

jrandom authored 20 years ago and zzz committed 20 years ago

two new tests - inactivity (seeing how we stay alive) and timeout (contacting someone who doesnt exist)

jrandom authored 20 years ago and zzz committed 20 years ago

  can respond apropriately
* optional inactivty timer with three possible results:
  disconnect, send a (blank) message (to be ACKed), or do nothing.

jrandom authored 20 years ago and zzz committed 20 years ago

* rework the messageInputStream to implement read(byte[], off, len), and fix some fencepost
  bugs in the byte retrieval

jrandom authored 20 years ago and zzz committed 20 years ago

    * Remove spurious flush calls from I2PTunnel, and work with the
      I2PSocket's output stream directly (as the various implementations
      do their own buffering).
    * Another pass at a long standing JobQueue bug - dramatically simplify
      the job management synchronization since we dont need to deal with
      high contention (unlike last year when we had dozens of queue runners
      going at once).
    * Logging

jrandom authored 20 years ago and zzz committed 20 years ago

    * Make the SAM bridge more resiliant to bad handshakes (thanks duck!)

jrandom authored 20 years ago and zzz committed 20 years ago

they may have restarted or otherwise lost the tags delivered to them.