5fca7ea49e
Back to VividCortex/ewma
2022-04-05 14:04:26 +02:00
66f019d886
Revert "regression: fix ewma warmup again ( #2079 )"
...
This reverts commit f67e9cab32
2022-04-03 23:01:03 +02:00
f67e9cab32
regression: fix ewma warmup again ( #2079 )
...
* Update estimators.go
* Update go.mod
* Update modules.txt
* Update go.sum
* Update serversInfo.go
* Update estimators.go
* Update serversInfo.go
2022-04-02 17:41:36 +02:00
df3fb0c9f8
Keep lines short
...
$ golines -w -m 120 --shorten-comments .
2022-03-23 17:48:48 +01:00
ca076ce133
Size estimator: provide the slide size
2022-02-21 14:16:13 +01:00
034d3bd424
Switch to lifenjoiner's ewma variant
2022-02-21 09:14:24 +01:00
9373cc7162
Use SimpleEWMA for the question size estimator
2022-02-20 23:40:32 +01:00
7956ba5b10
Switch to an ewma fork that allows setting the warmup samples #
2022-02-20 23:38:06 +01:00
b76db70a6c
Revert "cleanup: estimators: simplify blindAdjust"
...
This reverts commit c699e7bec4
2019-11-05 01:16:22 +01:00
c699e7bec4
cleanup: estimators: simplify blindAdjust
2019-10-18 20:24:11 +02:00
fa22cc32d7
Basic load balancing/failover
...
Try to send queries to one of the two fastest servers
2018-01-10 16:42:14 +01:00
efd0477c2b
Implement an actual estimator for the response size
...
Scale back the minimum question size when relevant.
Did I mention that this is yet another thing that was never properly
implemented in dnscrypt-proxy 1.x?
2018-01-10 09:46:27 +01:00
d8f8d561c8
Synthesize a truncated response if the response wouldn't fit the local MSS
2018-01-10 02:52:09 +01:00
841bf65d61
Reorganize
2018-01-09 18:32:14 +01:00
Frank Denis
livingentity
Markus Linnala