Timeline
11/18/06:
- 13:59 Changeset [2342] by
- thesis: Fixed the NF edge ext state machine diagram.
- 13:47 Changeset [2341] by
- thesis: Fixed the NF non-edge REA state machine diagram.
- 13:12 Changeset [2340] by
- thesis: Layout cleanups in the NF state machine diagram.
- 11:46 Changeset [2339] by
- thesis: Improved the layout of the NR diagram.
- 11:38 Changeset [2338] by
- Split the ni_session diagram into two parts.
11/17/06:
- 09:49 Changeset [2337] by
- Increased some more lifetimes in the configuration.
- 09:12 Changeset [2336] by
- Increased the maximum allowed session lifetime to 1000 seconds. Made the …
11/16/06:
- 18:14 Changeset [2335] by
- - debugging aid: enabled additional logging of internal message numbers
- 17:19 Changeset [2334] by
- - corrected timer output when retransmitting
- 15:33 Changeset [2333] by
- thesis: Two more diagrams for the eval chapter (both have to be improved).
- 15:32 Changeset [2332] by
- eval: dump_gnuplot.py now dumps results in microseconds.
- 14:09 Changeset [2331] by
- thesis: Added a diagram showing the differences between NAT and Firewall …
- 13:15 Changeset [2330] by
- thesis: Interpretation of the NF and NR benchmarks.
- 10:56 Changeset [2329] by
- thesis: More text and a diagram for the evaluation chapter.
- 00:55 Changeset [2328] by
- - removed retry_count and replaced it by retry_limit as defined in GIST …
11/15/06:
- 16:45 Changeset [2327] by
- - retry_count should be 7, according to T2=64000ms as upper bound for …
- 15:44 Changeset [2326] by
- thesis: Fixed several typos and smaller cosmetic issues.
- 12:43 Changeset [2325] by
- thesis: More docs on the session class. Dispatcher-Thread -> …
- 12:05 Changeset [2324] by
- thesis: Moved the test suite section to the evaluation chapter. Rewrote …
- 10:11 Changeset [2323] by
- thesis: Improved the generic state machine diagram explanations.
- 09:25 Changeset [2322] by
- Stress Test: Use the same session-ID for all CREATEs for now.
11/14/06:
- 23:36 Changeset [2321] by
- - client library implementation, compiled with sctplib v1.0.5
- 23:30 Changeset [2320] by
- - client library implementation, compiled with sctplib v1.0.5
- 16:53 Changeset [2319] by
- thesis: New layout for ni_session and the non-edge-ext state machine …
- 15:56 Changeset [2318] by
- thesis: fixed the NF-EDGE-REA state machine diagram (now EXT).
- 15:46 Changeset [2317] by
- thesis: Fixed the nf_session state machine diagram.
- 15:13 Changeset [2316] by
- thesis: Fixed the nr_session diagram.
- 14:11 Changeset [2315] by
- thesis: Improved the architecture diagram.
- 13:19 Changeset [2314] by
- thesis: Corrected the session overview diagram.
- 13:13 Changeset [2313] by
- thesis: Corrected ni_session diagram.
- 11:29 Changeset [2312] by
- thesis: First round of rewrite, based on Roland Bless' feedback.
- 09:40 Changeset [2311] by
- thesis: Added an interpretation of the NI benchmarks.
11/13/06:
- 15:25 Changeset [2310] by
- thesis: More benchmark interpretation.
- 14:30 Changeset [2309] by
- thesis: Added benchmark tables and documented some optimization ideas.
- 10:22 Changeset [2308] by
- thesis: Documented the methods used for measuring NATFW's performance.
11/10/06:
- 18:08 Changeset [2307] by
- Added an evaluation script that can dump both ASCII stats and a LaTeX …
- 15:58 Changeset [2306] by
- eval: Added a script which prints table rows in LaTeX syntax.
- 03:36 Changeset [2305] by
- - fixed sii_counter wrap
- 03:27 Changeset [2304] by
- After fixing some bugs (handshake works now), before changing SII handle …
11/09/06:
- 12:33 Changeset [2303] by
- eval: Forgot to restart the journal after the preparation phase in nf2.
- 12:09 Changeset [2302] by
- eval: Implemented a second NF test case. This one focuses on returning …
- 11:08 Changeset [2301] by
- Added a python script for generating gnuplot data files from a journal.
- 10:10 Changeset [2300] by
- eval: nf is now configurable and works in NAT mode, too.
- 03:03 Changeset [2299] by
- *** empty log message ***
11/08/06:
- 19:52 Changeset [2298] by
- - fixed indentation once again
- 13:51 Changeset [2297] by
- eval: Added an NR test case.
- 12:56 Changeset [2296] by
- Moved the evaluation python script to the eval directory. Made the …
- 12:34 Changeset [2295] by
- Increased the session_manager's table size to a large initial value. That …
- 12:09 Changeset [2294] by
- Moved the journal global variable to benchmark_journal.cpp Increased the …
- 10:44 Changeset [2293] by
- eval: Fixed the nop_dispatcher to still serialize messages.
- 10:15 Changeset [2292] by
- Make logging configurable in the eval code (via TEST_LOG environment …
- 10:10 Changeset [2291] by
- Added an eval test case for an NF node. Minor bug fixes in the existing …
- 09:39 Changeset [2290] by
- Added a method to the eval utils to create a valid ntlp APIMsg.
- 09:08 Changeset [2289] by
- Added a restart() method to the benchmark_journal class.
11/07/06:
- 22:56 Changeset [2288] by
- - added: send query in to_refresh_qnode() this was obviously forgotten. …
- 20:01 Changeset [2287] by
- - added NEED FIX comment
- 19:27 Changeset [2286] by
- - hopefully fixed query cookie code in send_response(): re-use query …
- 15:03 Changeset [2285] by
- Completed the eval/ni test program.
- 12:52 Changeset [2284] by
- Two more measuring points. Increased the default journal size. Display …
- 12:33 Changeset [2283] by
- Only write the benchmark journal on exit. Several smaller fixes, …
- 11:46 Changeset [2282] by
- Thesis: Added a page describing the static benchmarks.
- 10:29 Changeset [2281] by
- Renamed the NI eval test case.
- 10:07 Changeset [2280] by
- Implemented IEManager in a more efficient, hash_map based way. This …
- 05:16 Changeset [2279] by
- - hopefully fixed that an incoming query was ignored in state …
- 05:13 Changeset [2278] by
- - use of initializers
11/06/06:
- 11:11 Changeset [2277] by
- Added a program for testing an NF's throughput.
- 09:03 Changeset [2276] by
- Documented a gnu_cxx::hash_map vs. std::map issue. Fixed test suite …
11/04/06:
- 19:14 Changeset [2275] by
- - changed blocking behavior for recv
- 10:14 Changeset [2274] by
- Added an eval subdirectory for benchmarking parts of the natfw …
- 08:59 Changeset [2273] by
- Renamed the benchmark class to benchmark_journal. We will need the name …
11/02/06:
- 23:26 Changeset [2272] by
- - fixed sending error object for Invalid Cookies - using constants instead …
- 22:59 Changeset [2271] by
- - replaced several literal 255s with prot_query_encap constant
- 22:43 Changeset [2270] by
- - replaced several literal 255s with prot_query_encap constant
- 14:14 Changeset [2269] by
- Changed type of TPoverTCP's retry count to uint32 to avoid casting on …
- 12:04 Changeset [2268] by
- Fixed a bug in the write_header() method. Keep stress test client alive …
- 10:22 Changeset [2267] by
- Fixed race condition and segfault in the statemodule: TimerMsg? objects …
10/31/06:
- 21:19 Changeset [2266] by
- - added function is_bogus_source() to check for bad source addresses, i.e. …
- 21:18 Changeset [2265] by
- - added function is_bogus_source() to check for bad source addresses, i.e. …
- 21:14 Changeset [2264] by
- - retry sending if send() failed temporarily
10/30/06:
- 13:07 Changeset [2263] by
- New measuring points for serialization. Re-numbered and renamed some of …
- 12:54 Changeset [2262] by
- Several updates to the test suite (all borrowed from ntlp): * New test …
- 11:06 Changeset [2261] by
- - setting r_entry->rs_validity_time should happen before …
- 10:40 Changeset [2260] by
- - renamed raw to qenc, replaced literal 255 with prot_qeury_encap
- 10:29 Changeset [2259] by
- - hopefully fixed capability matching for non-SCTP protocols when SCTP is …
- 01:36 Changeset [2258] by
- - added/corrected stuff to negotiate SCTP as reliable transport
- 01:35 Changeset [2257] by
- - using constants for protocol address instead of raw (query_encap) or 254 …
10/29/06:
- 22:04 Changeset [2256] by
- - fixed indentation
- 21:55 Changeset [2255] by
- - converted uint128 to C++ class and added operator== to it
- 21:53 Changeset [2254] by
- - added simple coding tests
- 20:59 Changeset [2253] by
- - first draft for NetMsg?
10/27/06:
- 14:59 Changeset [2252] by
- Added more measuring points for benchmarking. Renumbered the existing …
- 13:04 Changeset [2251] by
- The stress testing client is now configurable via the command line.
- 11:09 Changeset [2250] by
- Added a minimal client for stress testing.
10/26/06:
- 15:41 Changeset [2249] by
- Added a python script which calculates stats from a benchmark journal.
- 10:51 Changeset [2248] by
- Corrected malformed benchmark journal header.
- 10:33 Changeset [2247] by
- Added Makefile variables for logging, debugging, etc.
- 09:40 Changeset [2246] by
- Write a human-readable header to the benchmark journal.
- 08:39 Changeset [2245] by
- Added a README file for the test suite.
10/25/06:
- 23:34 Changeset [2244] by
- - not include string gives compilation error on suse 9.3
- 13:55 Changeset [2243] by
- Benchmarking fix: Don't write the journal more than once.
- 13:42 Changeset [2242] by
- Use GIST from trunk instead of an own branch. Fixed Makefiles to pass …
- 12:07 Changeset [2241] by
- Documentation for benchmark's constructor.
- 10:45 Changeset [2240] by
- Added documentation to the Makefiles. Optimization can now be enabled via …
- 10:30 Changeset [2239] by
- Added measuring points throughout the code. Use clock_gettime() for making …
10/24/06:
- 22:30 Changeset [2238] by
- - introduced new GIST_Error class in order to get rid of GIST …
- 22:24 Changeset [2237] by
- - removed any GIST related error stuff
10/23/06:
- 20:32 Changeset [2236] by
- - fixed paths
- 17:50 Changeset [2235] by
- Cleaned up the mri_pathcoupled class. Several backwards compatible …
10/21/06:
- 15:24 Changeset [2234] by
- Changed Makefile to selectively enable debugging, logging, and …
- 12:22 Changeset [2233] by
- Fixed protlib to compile with -D_NO_LOGGING again. Minor Makefile cleanup …
10/20/06:
- 16:54 Changeset [2232] by
- *** empty log message ***
- 12:38 Changeset [2231] by
- Added an initial version of a class for benchmarking.
10/19/06:
- 21:56 Changeset [2230] by
- - fixed another timer from 50ms to retryperiod (500ms)
- 21:44 Changeset [2229] by
- - renamed RSV_time to RS_validity_time (should not be confused with …
- 21:25 Changeset [2228] by
- - fixed makefile for SCTP linkage
- 21:25 Changeset [2227] by
- - fixed makefile for SCTP linkage
- 21:23 Changeset [2226] by
- - no CVS entries required
- 21:22 Changeset [2225] by
- - removed wrong property
- 02:01 Changeset [2224] by
- - fixed compilation bug due to different interface in tp_over_sctp
- 01:23 Changeset [2223] by
- - added SCTP stuff - added new configuration parameters for sctpport, …
- 01:18 Changeset [2222] by
- - new parameter for requesting a reply in sendhello
- 01:18 Changeset [2221] by
- - reply for sendhello can be requested now - better log output if a timer …
- 01:15 Changeset [2220] by
- - indentation - new parameters for sctp and MA-Hello reply requests
- 01:13 Changeset [2219] by
- - adaptation to the new TP interface
Note: See TracTimeline
for information about the timeline view.
