| Commit message (Collapse) | Author | Age | |
|---|---|---|---|
| * | Always display per-class logs, even if the feature is disabled.bw_mixup | Mike Perry | 2011-12-05 |
| | | | | | | I want to try to disable it and see if these become more even or not. | ||
| * | Improve and demote some noisy loglines. | Mike Perry | 2011-12-04 |
| | | |||
| * | Clean up the spec a tad. | Mike Perry | 2011-12-03 |
| | | |||
| * | Allow us to cap feedback via consensus param. | Mike Perry | 2011-12-03 |
| | | | | | Also demote a couple logs, and add a couple more. | ||
| * | Make the pid target bw actually converge. | Mike Perry | 2011-12-02 |
| | | | | | | This seems to only take a few seconds to do.. Expensive, but probably not prohibitively so? | ||
| * | Be more merciful. | Mike Perry | 2011-12-02 |
| | | | | | | 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. | Mike Perry | 2011-12-02 |
| | | | | | | | 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. | Mike Perry | 2011-12-02 |
| | | | | | | 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. | Mike Perry | 2011-12-02 |
| | | | | | Otherwise we might not detect CPU overload conditions in them. | ||
| * | Actually, we should be using min(circ_error,pid_error). | Mike Perry | 2011-12-02 |
| | | |||
| * | Directly use the circ_error if it is negative. | Mike Perry | 2011-12-01 |
| | | |||
| * | Fix variable shadow, and add some logs. | Mike Perry | 2011-12-01 |
| | | |||
| * | Change how circuit failure is used. | Mike Perry | 2011-12-01 |
| | | | | | Let's try computing a second error and take their sum. | ||
| * | Implement per-class setpoints. | Mike Perry | 2011-12-01 |
| | | | | | Governed by consensus param (off by default). | ||
| * | Alter default param values. | Mike Perry | 2011-12-01 |
| | | | | | | | These defaults should still be equivalent to Section 2 in the spec, but these values will keep the PID interim state from disappearing every time ides crashes and the consensus params disappear. | ||
| * | Fix an exception that can happen with very large bandwidths. | Mike Perry | 2011-12-01 |
| | | | | | | We have a cap+log warn below that we will hit in this case instead of crashing. | ||
| * | Use desc bw if consensus bw is missing for some log lines | Mike Perry | 2011-11-20 |
| | | |||
| * | Add some notifications for #4359/#2286. | Mike Perry | 2011-11-20 |
| | | |||
| * | Be consistent about Wgd application for Guard+Exit nodes. | Mike Perry | 2011-11-20 |
| | | | | | Also correct log line. | ||
| * | Display a total of previous consensus bw measured in INFO logs. | Mike Perry | 2011-11-20 |
| | | |||
| * | Output git revision. | Mike Perry | 2011-11-20 |
| | | |||
| * | Bug 2014: Provide url to backup bw server. | Mike Perry | 2011-11-20 |
| | | |||
| * | Bug #4330: Have snakeinspector attach files if they existsoat-attachments | Mike Perry | 2011-11-19 |
| | | |||
| * | Forgot a spec paragraph.. | Mike Perry | 2011-11-17 |
| | | |||
| * | Merge branch 'pid_control-rebased' | Mike Perry | 2011-11-17 |
| |\ | |||
| | * | Let's try turning FetchDirInfo*Early on by default. | Mike Perry | 2011-11-17 |
| | | | | | | | | | Memory-only sqlite might make them more bearable. | ||
| | * | Update spec to include consensus params. | Mike Perry | 2011-11-17 |
| | | | | | | | | | | | These are still subject to change, especially in terms of defaults. | ||
| | * | Allow desc bw to be used in feedback. | Mike Perry | 2011-11-17 |
| | | | | | | | | | | | It may be equivalent to NS bw, since the NS bw functions as an accumulator.. We may want to set T_i_decay to 0 for this code. | ||
| | * | Submodule. | Mike Perry | 2011-11-17 |
| | | | |||
| | * | Switch to memory-only sqlalchemy by default. | Mike Perry | 2011-11-17 |
| | | | |||
| | * | Add some comments about potential timeout issues. | Mike Perry | 2011-11-17 |
| | | | |||
| | * | Update scanner cutoffs and timeouts in config. | Mike Perry | 2011-11-17 |
| | | | |||
| | * | Update pythonpath to modern versions. | Mike Perry | 2011-11-17 |
| | | | |||
| | * | Prevent file size list from double-listing percentiles. | Mike Perry | 2011-11-17 |
| | | | | | | | | | | | If two different file sizes serve one percentile, go with the smaller file size. | ||
| | * | Differentiate measured and update timestamps. | Mike Perry | 2011-11-17 |
| | | | | | | | | | | | | | | | This helps us to track Guard node feedback/non-feedback better. We may also end up using this to determine T_d and possibly T_i. Also improve some log messages. | ||
| | * | Retain pid_error_sum and pid_delta between votes. | Mike Perry | 2011-11-17 |
| | | | | | | | | | | | pid_delta is stored for informational purposes only. However, missing pid_error_sum for Guard measurements w/o feedback was a bug. | ||
| | * | Move more parameters to consensus. | Mike Perry | 2011-11-17 |
| | | | |||
| | * | Fix for 0-bandwidth Guard issue. | Mike Perry | 2011-11-17 |
| | | | | | | | | | We were accidentally voting 0 for guards that weren't measured this round. | ||
| | * | Tweak a regex and add some log lines. | Mike Perry | 2011-11-14 |
| | | | |||
| | * | Tweaks to try to make PID look more like other bw auths. | Mike Perry | 2011-11-14 |
| | | | | | | | | | | | | | | | | | Just to make sure something crazy is not going on, let's try eliminating the circ failure dampening, and re-instate the "larger ratio" idea for PID control. With these both re-enabled, PID "feedback" should look more similar to the other bandwidth auths, because with just one auth running the code, there is no feedback. | ||
| | * | Bug #4268: Remove sleeping from bw auths | Mike Perry | 2011-11-14 |
| | | | | | | | | | | | Code is still there, just disabling via config, and adding a check to ensure the time isn't negative. | ||
| | * | Bug #4273: Reduce max fetch time | Mike Perry | 2011-11-14 |
| | | | | | | | | | | | 5 minutes seems better. Now that we automatically select slice file sizes to complete in about 5 seconds, it should be more than enough. | ||
| | * | Prior to feedback, we used the greater of two ratios. | Mike Perry | 2011-11-14 |
| | | | | | | | | | | | Let's keep this behavior in place when the bwauthpid consensus param is missing. | ||
| | * | Bugs 4431+2014: Randomized url support + variable file sizes. | Mike Perry | 2011-11-14 |
| | | | | | | | | | Scanner child piece. | ||
| | * | Bug #4431: Automatically determine size cutoffs. | Mike Perry | 2011-11-14 |
| | | | | | | | | | Also update the spec and the use of Wgd. | ||
| | * | Fix accidental reset of circ_fail_rate. | Mike Perry | 2011-11-14 |
| | | | |||
| | * | Submodule. | Mike Perry | 2011-11-14 |
| | | | |||
| | * | Actually remove stale files. | Mike Perry | 2011-11-14 |
| | | | | | | | | | Also kill some dead code. | ||
| | * | Update cron script to differentiate output vs archive. | Mike Perry | 2011-11-14 |
| | | | |||
| | * | Handle negative bandwidths. | Mike Perry | 2011-11-14 |
| | | | | | | | | | Also record the circ fail rate. | ||
 |
index : user/mikeperry/torflow | |
| Mike's TorFlow repository | Mike Perry |
| summaryrefslogtreecommitdiff |