* Take out the router watchdog's teeth (don't restart on leaseset failure)
* Filter the IRC ping/pong messages, as some clients send unsafe
information in them (thanks aardvax and dust!)
* Major SSU and router tuning to reduce contention, memory usage, and GC
churn. There are still issues to be worked out, but this should be a
substantial improvement.
* Modified the optional netDb harvester task to support choosing whether
to use (non-anonymous) direct connections or (anonymous) exploratory
tunnels to do the harvesting. Harvesting itself is enabled via the
advanced config "netDb.shouldHarvest=true" (default is false) and the
connection type can be chosen via "netDb.harvestDirectly=false" (default
is false).
* Bugfix to the I2PTunnel web config to properly accept i2cp port settings
* Initial sucker refactoring to simplify reuse of the html parsing
* Beginnings of hooks to push imported rss/atom out to remote syndie
archives automatically (though not enabled currently)
* Further SSU peer test cleanup
* Run the peer profile coalescing/reorganization outside the job queue
(on one of the timers), to cut down on some job queue congestion. Also,
trim old profiles while running, not just when starting up.
* Slightly more sane intra-floodfill-node netDb activity (only flood new
entries)
* Workaround in the I2PTunnelHTTPServer for some bad requests (though the
source of the bug is not yet addressed)
* Better I2PSnark reconnection handling
* Further cleanup in the new tunnel build process
* Make sure we expire old participants properly
* Remove much of the transient overload throttling (it wasn't using a good
metric)
* Rather than profile individual tunnels for throughput over their
lifetime, do so at 1 minute intervals (allowing less frequently active
tunnels to be more fairly measured).
* Run the live tunnel load test across two tunnels at a time, by default.
The load test runs for a random period from 90s to the tunnel lifetime,
self paced. This should help gathering data for profiling peers that
are in exploratory tunnels.
2006-01-03 jrandom
* Calculate the overall peer throughput across the 3 fastest one minute
tunnel throughput values, rather than the single fastest throughput.
* Degrade the profiled throughput data over time (cutting the profiled
peaks in half once a day, on average)
* Enable yet another new speed calculation for profiling peers, using the
peak throughput from individual tunnels that a peer is participating in,
rather than across all tunnels they are participating in. This helps
gather a fairer peer throughput measurement, since it won't allow a slow
high capacity peer seem to have a higher throughput (pushing a little
data across many tunnels at once, as opposed to lots of data across a
single tunnel). This degrades over time like the other.
* Add basic OS/2 support to the jbigi code (though we do not bundle a
precompiled OS/2 library)
* Added some I2PSnark sanity checks, an OOMListener when running
standalone, and a guard against keeping memory tied up indefinitely.
* Sanity check on the watchdog (thanks zzz!)
* Handle invalid HTTP requests in I2PTunnel a little better
* IE doesn't strip SPAN from <button> form fields, so add in a workaround
within I2PTunnel.
* Increase the maximum SSU retransmission timeout to accomodate slower or
more congested links (though SSU's RTO calculation will usually use a
much lower timeout)
* Moved the streaming lib timed events off the main timer queues and onto
a streaming lib specific set of timer queues. Streaming lib timed
events are more likely to have lock contention on the I2CP socket while
other timed events in the router are (largely) independent.
* Fixed a case sensitive lookup bug (thanks tino!)
* Syndie cleanup - new edit form on the preview page, and fixed some blog
links (thanks tino!)
* More aggressive I2PTunnel content encoding munging to work around some
rare HTTP behavior (ignoring q values on Accept-encoding, using gzip
even when only identity is specified, etc). I2PTunnelHTTPServer now
sends "Accept-encoding: \r\n" plus "X-Accept-encoding: x-i2p-gzip\r\n",
and I2PTunnelHTTPServer handles x-i2p-gzip in either the Accept-encoding
or X-Accept-encoding headers. Eepsite operators who do not know to
check for X-Accept-encoding will simply use the identity encoding.
* Initial pass of the routerconsole revamp, starting with I2PTunnel and
being progressively rolled out to other sections at later dates.
Featuring abstracted W3C strict XHTML1.0 markup, with CSS providing
layout and styling.
* Implemented console themes. Users can create their own themes by
creating css files in: {i2pdir}/docs/themes/console/{themename}/
and activating it using the routerconsole.theme={themename} advanced
config property. Look at the example incomplete "defCon1" theme.
Note: This is very much a work in progress. Folks might want to hold-off
creating their own skins until the markup has solidified.
* Added "routerconsole.javascript.disabled=true" to disable console
client-side scripting and "routerconsole.css.disabled=true" to remove
css styling (only rolled out in the i2ptunnel interface currently)
* Fixed long standing bug with i2ptunnel client and server edit screens
where tunnel count and depth properties would fail to save. Added
backup quantity and variance configuration options.
* Added basic accessibility support (key shortcuts, linear markup, alt and
title information and form labels).
* So far only tested on IE6, Firefox 1.0.6, Opera 8 and lynx.
* In Syndie, propogate the subject and tags in a reply, and show the parent
post on the edit page for easy quoting. (thanks identiguy and CofE!)
* Streamline some netDb query handling to run outside the jobqueue -
which means they'll run on the particular SSU thread that handles the
message. This should help out heavily loaded netDb peers.
* Fix bug in ircclient that prevented it to use its own dest (i.e. was
always shared. (thx for info Ragnarok)
* Fix crash in Sucker with some bad html.
* Choke SSU retransmissions to a peer while there is already a
retransmission in flight to them. This currently lets other initial
transmissions through, since packet loss is often sporadic, but maybe
this should block initial transmissions as well?
* Display the retransmission bytes stat on peers.jsp (thanks bar!)
* Filter QUIT messages in the I2PTunnelIRCClient proxy
* Implemented a new I2PTunnelIRCClient which locally filters inbound and
outbound IRC commands for anonymity and security purposes, removing all
CTCP messages except ACTION, as well as stripping the hostname from the
USER message (while leaving the nick and 'full name'). The IRC proxy
doesn't use this by default, but you can enable it by creating a new
"IRC proxy" tunnel on the web interface, or by changing the tunnel type
to "ircclient" in i2ptunnel.config.
2005-10-10 jrandom
* I2PTunnel http client config cleanup and stats
* Minor SSU congestion tweaks and stats
* Reduced netDb exploration period
* Allow the I2PTunnelHTTPServer to send back the first few packets of an
HTTP response quicker, and initialize the streaming lib's cwin more
carefully.
* Added a small web UI to the new Syndie scheduled updater. If you log in
as a user authorized to use the remote archive funtionality, you can
request remote archives in your address book to be automatically pulled
down by checking the "scheduled?" checkbox.
* I2PTunnel bugfix (thanks Complication!)
* Increase the SSU cwin slower during congestion avoidance (at k/cwin^2
instead of k/cwin)
* Limit the number of inbound SSU sessions being built at once (using
half of the i2np.udp.maxConcurrentEstablish config prop)
* Don't shitlist on a message send failure alone (unless there aren't any
common transports).
* More careful bandwidth bursting
* Reworded the SSU introductions config section (thanks duck!)
* Force identity content encoding for I2PTunnel httpserver requests
(thanks redzara!)
* Further x-i2p-gzip bugfixes for the end of streams
* Reduce the minimum bandwidth limits to 3KBps steady and burst (though
I2P's performance at 3KBps is another issue)
* Cleaned up some streaming lib structures
* Allow reseeding on the console if the netDb knows less than 30 peers,
rather than less than 10 (without internet connectivity, we keep the
last 15 router references)
* Reenable the x-i2p-gzip HTTP processing by default, flushing the stream
more aggressively.
* Show the status that used to be called "ERR-Reject" as "OK (NAT)"
* Reduced the default maximum number of streaming lib resends of a packet
(10 retransmits is a bit much with a reasonable RTO)
* Better i2paddresshelper handling in the I2PTunnel httpclient, plus a new
conflict resolution page if the i2paddresshelper parameter differs from
an existing name to destination mapping.
2005-09-25 jrandom
* Fix a long standing streaming lib bug (in the inactivity detection code)
* Improved handling of initial streaming lib packet retransmissions to
kill the "lost first packet" bug (where a page shows up with the first
few KB missing)
* Add support for initial window sizes greater than 1 - useful for
eepsites to transmit e.g. 4 packets full of data along with the initial
ACK, thereby cutting down on the rtt latency. The congestion window
size can and does still shrink down to 1 packet though.
* Adjusted the streaming lib retransmission calculation algorithm to be
more TCP-like.
* Added the natively compiled jbigi and patched java service wrapper for
OS X. Thanks Bill Dorsey for letting me use your machine!
* Don't build i2p.exe or i2pinstall.exe when run on OS X machines, as we
don't bundle the binutils necessary (and there'd be a naming conflict
if we did).
* Added 'single user' functionality to syndie - if the single user
checkbox on the admin page is checked, all users are allowed to control
the instance and sync up with remote syndie nodes.
* Temporarily disable the x-i2p-gzip in i2ptunnel until it is more closely
debugged.
* Adjust I2PTunnelHTTPServer so it can be used for outproxy operators
(just specify the spoofed host as an empty string), allowing them to
honor x-i2p-gzip encoding.
* Let windows users build the exes too (thanks bar and redzara!)
* Allow I2PTunnel httpserver operators to disable gzip compression on
individual tunnels with the i2ptunnel.gzip=false client option
(good idea susi!)
* Added the i2p.exe and i2pinstall.exe for windows users, using launch4j.
* Added runplain.sh for *nix/osx users having problems using the java
service wrapper (called from the install dir as: sh runplain.sh)
* Bundle susidns and syndie, with links on the top nav
* Have I2PTunnelHTTPClient and I2PTunnelHTTPServer use the x-i2p-gzip
content-encoding (if offered), reducing the payload size before it
reaches the streaming lib. The existing compression is at the i2cp
level, so we've been packetizing 4KB of uncompressed data and then
compressing those messages, rather than compressing and then packetizing
4KB of compressed data. This should reduce the number of round trips
to fetch web pages substantially.
* Adjust the startup and timing of the addressbook so that susidns always
has config to work off, and expose a method for susidns to tell it to
reload its config and rerun.
* Don't persist peer profiles until we are shutting down, as the
persistence process gobbles RAM and wall time.
* Bugfix to allow you to check/uncheck the sharedClient setting on the
I2PTunnel web interface.
* Be more careful when expiring a failed tunnel message fragment so we
don't drop the data while attempting to read it.
* Don't publish leaseSets to the netDb if they will never be looked for -
namely, if they are for destinations that only establish outbound
streams. I2PTunnel's 'client' and 'httpclient' proxies have been
modified to tell the router that it doesn't need to publish their
leaseSet (by setting the I2CP config option 'i2cp.dontPublishLeaseSet'
to 'true').
* Don't publish the top 10 peer rankings of each router in the netdb, as
it isn't being watched right now.
* Minor logging and optimization tweaks in the router and SDK
* Use ISO-8859-1 in the XML files (thanks redzara!)
* The consolePassword config property can now be used to bypass the router
console's nonce checking, allowing CLI restarts
* Catch errors with corrupt tunnel messages more gracefully (no need to
kill the thread and cause an OOM...)
* Don't skip shitlisted peers for netDb store messages, as they aren't
necessarily shitlisted by other people (though they probably are).
* Adjust the netDb store per-peer timeout based on each particular peer's
profile (timeout = 4x their average netDb store response time)
* Don't republish leaseSets to *failed* peers - send them to peers who
replied but just didn't know the value.
* Set a 5 second timeout on the I2PTunnelHTTPServer reading the client's
HTTP headers, rather than blocking indefinitely. HTTP headers should be
sent entirely within the first streaming packet anyway, so this won't be
a problem.
* Don't use the I2PTunnel*Server handler thread pool by default, as it may
prevent any clients from accessing the server if the handlers get
blocked by the streaming lib or other issues.
* Don't overwrite a known status (OK/ERR-Reject/ERR-SymmetricNAT) with
Unknown.
2005-07-27 jrandom
* Enabled SSU as the default top priority transport, adjusting the
config.jsp page accordingly.
* Add verification fields to the SSU and TCP connection negotiation (not
compatible with previous builds)
* Enable the backwards incompatible tunnel crypto change as documented in
tunnel-alt.html (have each hop encrypt the received IV before using it,
then encrypt it again before sending it on)
* Disable the I2CP encryption, leaving in place the end to end garlic
encryption (another backwards incompatible change)
* Adjust the protocol versions on the TCP and SSU transports so that they
won't talk to older routers.
* Fix up the config stats handling again
* Fix a rare off-by-one in the SSU fragmentation
* Reduce some unnecessary netDb resending by inluding the peers queried
successfully in the store redundancy count.
* Reduced the growth factor on the slow start and congestion avoidance for
the streaming lib.
* Adjusted some of the I2PTunnelServer threading to use a small pool of
handlers, rather than launching off new threads which then immediately
launch off an I2PTunnelRunner instance (which launches 3 more threads..)
* Don't persist session keys / session tags (not worth it, for now)
* Added some detection and handling code for duplicate session tags being
delivered (root cause still not addressed)
* Make the PRNG's buffer size configurable (via the config property
"i2p.prng.totalBufferSizeKB=4096")
* Disable SSU flooding by default (duh)
* Updates to the StreamSink apps for better throttling tests.
* Within the tunnel, use xor(IV, msg[0:16]) as the flag to detect dups,
rather than the IV by itself, preventing an attack that would let
colluding internal adversaries tag a message to determine that they are
in the same tunnel. Thanks dvorak for the catch!
* Drop long inactive profiles on startup and shutdown
* /configstats.jsp: web interface to pick what stats to log
* Deliver more session tags to account for wider window sizes
* Cache some intermediate values in our HMACSHA256 and BC's HMAC
* Track the client send rate (stream.sendBps and client.sendBpsRaw)
* UrlLauncher: adjust the browser selection order
* I2PAppContext: hooks for dummy HMACSHA256 and a weak PRNG
* StreamSinkClient: add support for sending an unlimited amount of data
* Migrate the tests out of the default build jars
2005-06-22 Comwiz
* Migrate the core tests to junit
* More fixes for the I2PTunnel "other" interface handling (thanks nelgin!)
* Add back the code to handle bids from multiple transports (though there
is still only one transport enabled by default)
* Adjust the router's queueing of outbound client messages when under
heavy load by running the preparatory job in the client's I2CP handler
thread, thereby blocking additional outbound messages when the router is
hosed.
* No need to validate or persist a netDb entry if we already have it
And for some udp stuff:
* only bid on what we know (duh)
* reduceed the queue size in the UDPSender itself, so that ACKs go
through more quickly, leaving the payload messages to queue up in
the outbound fragment scheduler
* rather than /= 2 on congestion, /= 2/3 (still AIMD, but less drastic)
* adjust the fragment selector so a wsiz throttle won't force extra
volleys
* mark congestion when it occurs, not after the message has been
ACKed
* when doing a round robin over the active messages, move on to the
next after a full volley, not after each packet (causing less "fair"
performance but better latency)
* reduced the lock contention in the inboundMessageFragments by
moving the ack and complete queues to the ACKSender and
MessageReceiver respectively (each of which have their own
threads)
* prefer new and existing UDP sessions to new TCP sessions, but
prefer existing TCP sessions to new UDP sessions
* In the SDK, we don't actually need to block when we're sending a message
as BestEffort (and these days, we're always sending BestEffort).
* Pass out client messages in fewer (larger) steps.
* Have the InNetMessagePool short circuit dispatch requests.
* Have the message validator take into account expiration to cut down on
false positives at high transfer rates.
* Allow configuration of the probabalistic window size growth rate in the
streaming lib's slow start and congestion avoidance phases, and default
them to a more conservative value (2), rather than the previous value
(1).
* Reduce the ack delay in the streaming lib to 500ms
* Honor choke requests in the streaming lib (only affects those getting
insanely high transfer rates)
* Let the user specify an interface besides 127.0.0.1 or 0.0.0.0 on the
I2PTunnel client page (thanks maestro^!)
(plus minor udp tweaks)
