 evgeny
|
34e3d9bba0
|
Verify queue budget saturation and the BBR fast path contract
|
2026-10-23 01:02:15 +02:00 |
|
|
evgeny
|
599d8ad421
|
Bound BBR probe windows by the steady-state queue time budget
|
2026-10-23 01:01:25 +02:00 |
|
|
evgeny
|
4035794631
|
Define and test the BBR steady-state queue time budget
|
2026-10-23 01:00:36 +02:00 |
|
|
evgeny
|
5f406c2a41
|
Document BBR queue budget conflict with the fixed latency target
|
2026-10-23 00:58:29 +02:00 |
|
|
evgeny
|
94d6dd05c8
|
Expose BBR bandwidth and ACK budget alongside physical queue peaks
|
2026-10-23 00:54:27 +02:00 |
|
|
evgeny
|
67ae4ecf11
|
Record full regression results and reproducible NAT profile failures
|
2026-10-23 00:51:44 +02:00 |
|
|
evgeny
|
d0f122f4d8
|
Record verified RTT probe fixes and remaining multipath failures
|
2026-10-23 00:49:02 +02:00 |
|
|
evgeny
|
08b6b2e2a1
|
Document DATA MSS and shared BBR transport window contracts
|
2026-10-23 00:45:09 +02:00 |
|
|
evgeny
|
5298fe9f6a
|
Keep BBR loss bounds above the DATA window minimum
|
2026-10-23 00:43:54 +02:00 |
|
|
evgeny
|
aefa8ff680
|
Check BBR loss bounds against the minimum DATA window
|
2026-10-23 00:43:31 +02:00 |
|
|
evgeny
|
0f23076f48
|
Apply BBR DATA windows consistently during startup and RTT probes
|
2026-10-23 00:42:45 +02:00 |
|
|
evgeny
|
92a535e009
|
Cover BBR probe drain and transport DATA window consistency
|
2026-10-23 00:42:02 +02:00 |
|
|
evgeny
|
519d6e4eaf
|
Record verified BBR fixes and remaining multipath contract conflicts
|
2026-10-23 00:25:55 +02:00 |
|
|
evgeny
|
716db3a820
|
Saturate BBR byte windows instead of wrapping at large BDP
|
2026-10-23 00:24:35 +02:00 |
|
|
evgeny
|
3d6f855d34
|
Keep BBR probing and aggregation counters dimensionally consistent in bytes
|
2026-10-23 00:17:14 +02:00 |
|
|
evgeny
|
229e992695
|
Cover probing byte budgets and ACK aggregation beyond 64KiB
|
2026-10-23 00:16:40 +02:00 |
|
|
evgeny
|
a79c724110
|
Grow BBR probe inflight bounds in MSS-sized byte increments
|
2026-10-23 00:15:46 +02:00 |
|
|
evgeny
|
6b1298aa62
|
Reproduce byte-sized instead of MSS-sized BBR probing growth
|
2026-10-23 00:15:28 +02:00 |
|
|
evgeny
|
0364abb00e
|
Log sampled BBR model state without flooding transport diagnostics
|
2026-10-23 00:12:28 +02:00 |
|
|
evgeny
|
f8d532b04d
|
Describe ACK handling and bounded DATA pacing in the ETCP protocol
|
2026-10-23 00:10:01 +02:00 |
|
|
evgeny
|
4dba742149
|
Document BBR byte units and flight-based delivery rate samples
|
2026-10-23 00:10:00 +02:00 |
|
|
evgeny
|
88eb0f2108
|
Require multipath goodput without inflating channel queues
|
2026-10-23 00:08:15 +02:00 |
|
|
evgeny
|
ade0e93f7f
|
Measure actual channel queueing delay independently of propagation RTT
|
2026-10-23 00:08:15 +02:00 |
|
|
evgeny
|
addff8547e
|
Keep the ACK compression fixture within its known retransmission deadline
|
2026-10-23 00:06:30 +02:00 |
|
|
evgeny
|
1c18bdc9ab
|
Measure BBR delivery over packet flights instead of inter-ACK gaps
|
2026-10-23 00:06:30 +02:00 |
|
|
evgeny
|
f3008b7aba
|
Snapshot delivery and send intervals for per-link BBR rate measurement
|
2026-10-23 00:06:30 +02:00 |
|
|
evgeny
|
c2f4f048d7
|
Reproduce inflated delivery-rate samples when ACKs arrive together
|
2026-10-23 00:05:48 +02:00 |
|
|
evgeny
|
3c38072b9b
|
Send ACK independently of the local outgoing DATA congestion pacer
|
2026-10-23 00:03:29 +02:00 |
|
|
evgeny
|
4cfd0ec89f
|
Identify DATA in local datagram metadata without changing the wire header
|
2026-10-23 00:03:29 +02:00 |
|
|
evgeny
|
08c42e62ef
|
Reproduce ACK starvation behind the reverse-direction DATA pacer
|
2026-10-23 00:02:57 +02:00 |
|
|
evgeny
|
dd153264e1
|
Verify byte-rate scaling at multiple MTUs and multi-gigabit rates
|
2026-10-23 00:00:40 +02:00 |
|
|
evgeny
|
6f87ea1426
|
Keep BBR pacing and BDP in byte units without a second MSS factor
|
2026-10-22 23:59:57 +02:00 |
|
|
evgeny
|
6fec5d6d98
|
Reproduce MSS-dependent BBR pacing and BDP with a dimensional regression
|
2026-10-22 23:59:20 +02:00 |
|
|
evgeny
|
9b9a9f2e00
|
Record final multipath fault results and sanitizer validation
|
2026-10-22 23:41:45 +02:00 |
|
|
evgeny
|
27a7b7bf25
|
Declare the multipath phony target without overlapping Automake conditions
|
2026-10-22 23:40:12 +02:00 |
|
|
evgeny
|
c9ba018a0d
|
Bound fault steady goodput by newly received unique DATA
|
2026-10-22 23:37:24 +02:00 |
|
|
evgeny
|
df5f6263bd
|
Measure short HOLD across resume and arm faults before propagation completes
|
2026-10-22 23:37:02 +02:00 |
|
|
evgeny
|
0b5a892a88
|
Document multipath acceptance criteria and reproduced protocol limitations
|
2026-10-22 23:36:21 +02:00 |
|
|
evgeny
|
076016f34a
|
Expose multipath queue accounting and add a dedicated long-running check target
|
2026-10-22 23:34:20 +02:00 |
|
|
evgeny
|
8417b1955c
|
Document emulator queue bounds and account for allocation failures
|
2026-10-22 23:33:14 +02:00 |
|
|
evgeny
|
67322582f6
|
Report delivery failures per scenario without hiding subsequent channel cases
|
2026-10-22 23:32:34 +02:00 |
|
|
evgeny
|
439361cb52
|
Exclude previously buffered data from multipath utilization results
|
2026-10-22 23:31:02 +02:00 |
|
|
evgeny
|
6e350760d1
|
Check active channel stalls, delivery deadlines and delayed packet recovery
|
2026-10-22 23:29:29 +02:00 |
|
|
evgeny
|
672f04b833
|
Exercise real ETCP multipath against independent bandwidth and RTT profiles
|
2026-10-22 23:23:05 +02:00 |
|
|
evgeny
|
f5cab5a14e
|
Calibrate dummynet bandwidth and verify stalled channel lifecycle
|
2026-10-22 23:19:01 +02:00 |
|
|
evgeny
|
f5f6ff50af
|
Make dummynet bandwidth timing cumulative and add bounded DROP/HOLD modes
|
2026-10-22 23:18:11 +02:00 |
|
|
evgeny
|
62f4850968
|
Record successful full ETCP regression validation
|
2026-10-22 23:04:00 +02:00 |
|
|
evgeny
|
082d9d3960
|
Record STCP SIGPIPE regression and sanitizer validation
|
2026-10-22 23:02:42 +02:00 |
|
|
evgeny
|
05a5b6c2a7
|
Exercise STCP writes with default SIGPIPE disposition
|
2026-10-22 22:59:45 +02:00 |
|
|
evgeny
|
fdcf92be04
|
Prevent SIGPIPE from terminating embedded STCP transports
|
2026-10-22 22:59:44 +02:00 |
|