user/mikeperry/torflow, branch logging Mike's TorFlow repository The old code should use circ failures.. 2011-12-06T00:18:06+00:00 Mike Perry mikeperry-git@fscked.org 2011-12-06T00:18:06+00:00 2416763b6351d5ad2e7514677412fc501e9dda4e I think this was added because circ_fail_rate used to get set to 0 by a consensus param that we later changed. 
I think this was added because circ_fail_rate used to get set to 0 by a
consensus param that we later changed.
Always display per-class logs, even if the feature is disabled. 2011-12-06T00:16:52+00:00 Mike Perry mikeperry-git@fscked.org 2011-12-05T23:31:18+00:00 0b5fae707b0dc98f45c6e8f81ff22cfb61d05069 I want to try to disable it and see if these become more even or not. 
I want to try to disable it and see if these become more even
or not.
Improve and demote some noisy loglines. 2011-12-04T22:49:05+00:00 Mike Perry mikeperry-git@fscked.org 2011-12-04T05:41:50+00:00 a0de14874f23873728320ba378c364676b9740dd 






Clean up the spec a tad.
2011-12-04T02:13:37+00:00

Mike Perry
mikeperry-git@fscked.org

2011-12-04T02:13:37+00:00

2d4b7ba817e6c398f5b0137d893d96957ac2adf7












Allow us to cap feedback via consensus param.
2011-12-04T00:53:44+00:00

Mike Perry
mikeperry-git@fscked.org

2011-12-04T00:53:44+00:00

0300422f1f9a1c9ab6df2344a343898ee545082f

Also demote a couple logs, and add a couple more.





Also demote a couple logs, and add a couple more.







Make the pid target bw actually converge.
2011-12-03T01:04:04+00:00

Mike Perry
mikeperry-git@fscked.org

2011-12-03T01:04:04+00:00

063ea9bca63343996a8e3e3f6d3c2142bad8681e

This seems to only take a few seconds to do.. Expensive, but probably not
prohibitively so?





This seems to only take a few seconds to do.. Expensive, but probably not
prohibitively so?







Be more merciful.
2011-12-03T00:35:09+00:00

Mike Perry
mikeperry-git@fscked.org

2011-12-03T00:35:09+00:00

0e2cb272706083f68ce4b7a8244403580ce37f1b

We can just check to see if we've previously punished the node by inspecting
the previous values directly. Forgot about that.





We can just check to see if we've previously punished the node by inspecting
the previous values directly. Forgot about that.







Implement mercy, subject to consensus parameter.
2011-12-03T00:18:59+00:00

Mike Perry
mikeperry-git@fscked.org

2011-12-03T00:17:16+00:00

0e0564672f92ef8332b0707e7ad5ddfa0e25dd68

If the param is set, we don't accumulate negative pid_error_sum for nodes with
negative pid_error. Prevents us from punishing these nodes to 0 bw. The param
only makes sense for descriptor-based feedback.





If the param is set, we don't accumulate negative pid_error_sum for nodes with
negative pid_error. Prevents us from punishing these nodes to 0 bw. The param
only makes sense for descriptor-based feedback.







Add a consensus param to raise the PID setpoint.
2011-12-03T00:18:56+00:00

Mike Perry
mikeperry-git@fscked.org

2011-12-02T23:58:58+00:00

65b9670fc3db3708c4f069ce27dd06a788926415

Intuitively, it seems to make sense to exclude nodes with no chance
of reaching the setpoint from consideration.





Intuitively, it seems to make sense to exclude nodes with no chance
of reaching the setpoint from consideration.







Use FastFirstHopPK to get circ fail rates for Guards.
2011-12-02T22:16:00+00:00

Mike Perry
mikeperry-git@fscked.org

2011-12-02T22:16:00+00:00

8684b9588d60a05fed11c83380540c0a8abbf6ca

Otherwise we might not detect CPU overload conditions in them.





Otherwise we might not detect CPU overload conditions in them.