summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
* Fix README for the file server.logbumpMike Perry2015-06-01
|
* Demote loglevel back down to notice.Mike Perry2015-06-01
| | | | Also fix issues with logging system usage and upgrade some messages.
* Workaround strange issue with unmeasured flag parsing.Mike Perry2015-05-31
| | | | | | | | It seems like consensus/newdesc events are causing TorCtl to reset the unmeasured flag when we shouldn't be. That is a TorCtl bug, but since there may also be memory bloat issues with leaving the bwauthority_child processes around anyway, just exit and let the parent restart the slice with a clean consensus.
* Stop listening for consensus events once we have it.Mike Perry2015-05-31
|
* Double the scanners, double the torsTom Ritter2015-05-31
| | | | Also update aggregate.py to work with a two-tor system
* Fix .gitignoresTom Ritter2015-05-31
| | | | Make them actually ignore the data files that are generated, and prevent the .gitignores from being deleted by run_scan.sh
* We now support python2.6 or python2.7.Mike Perry2015-05-31
| | | | Adjust setup accordingly. Also bump other version deps.
* Ignore SSL Cert ErrorsTom Ritter2015-05-31
| | | | A point release in python turned on cert validation (yay). This broke the script (boo). Ignore cert validation, as the only point of SSL here is to prevent caching, not confidentiality.
* Update pathsTom Ritter2015-05-31
| | | | TorCtl is located in a different location if you actually look at the README and the shell script for the submodule. Tor is cloned without the '.git' by default, unless you name it that way, which is confusing. Dependencies have been updated (kinda.)
* Use tjr's bwfile size values as the initial defaults.Mike Perry2015-05-31
|
* Turn on code to add exits if few are present in a slice.Mike Perry2015-05-31
|
* Add a 5th scanner for Unmeasured=1 nodes.Mike Perry2015-05-31
|
* Update TorCtl submodule.Mike Perry2015-05-31
|
* Log the success counts for each slice.Mike Perry2015-05-26
|
* Create script to Use virtualenv to setup everything.Mike Perry2015-05-26
| | | | This will help us pin versions to eliminate bitrot issues.
* Add section to README.BwAuthorities with pip/peep instructions.Isis Lovecruft2015-05-20
|
* Add a pip/peep requirements.txt file for running a Bandwidth Authority.Isis Lovecruft2015-05-20
|
* Create scan-data directories if they do not exist.aagbsn2014-11-06
| | | | | | Fixes #13663. Thanks to micah for reporting this issue.
* Fixes #13645, removal of bwauthority.cfg causes bwauth to failaagbsn2014-11-04
| | | | Thanks to micah for catching and patching this bug :-)
* fix on local.cfg patch: Work with dash too and not just bashPeter Palfrader2014-10-13
|
* Only try to kill tor if it is our process and it existsPeter Palfrader2014-10-13
|
* Fix indentingPeter Palfrader2014-10-13
|
* Do not export an empty PYTHONPATHPeter Palfrader2014-10-13
|
* Load config from local.cfg file if it existsPeter Palfrader2014-10-13
|
* Fix a comment/output to match realityPeter Palfrader2014-10-13
|
* Make the find call sanerPeter Palfrader2014-10-13
| | | | | Fix the find call to not require grep, and to not exec up to a 1000 rms at once, or exec rms with no arguments.
* Make default bw auth loglevel WARN.Mike Perry2014-03-18
|
* Update TorCtl submodule for 9771aagbsn2013-09-20
|
* 9791 - Remove gap in BwAuthority coverageaagbsn2013-09-20
|
* Update TorCtl submodule for 9771aagbsn2013-09-20
|
* Remove use_unmeasured parameteraagbsn2013-04-14
|
* Update Submodule TorCtlaagbsn2013-03-14
|
* 8417 - Add bwfiles fileaagbsn2013-03-07
|
* Update Submodule TorCtl for 8273 supportaagbsn2013-03-01
|
* 8273 Measure relays with Unmeasured flag by defaultaagbsn2013-03-01
|
* Update submodule TorCtlaagbsn2013-03-01
|
* Update README.BwAuthorities with TorCtl instructionsaagbsn2013-01-09
|
* Use https, not unauthenticated git.aagbsn2013-01-09
| | | | Thanks to weasel for spotting this
* Clarify some datapoints.Mike Perry2012-10-11
| | | | I mean hey, maybe not everyone can rollerskate on multidementianal surfaces..
* Clarify that trials are circuit counts in output.Mike Perry2012-10-10
|
* Update results file with new output.Mike Perry2012-10-10
|
* Add a Guard DoS simulation.Mike Perry2012-10-10
|
* Represent false positives as raw counts.Mike Perry2012-10-10
| | | | Also, clarify the startup false positive test.
* Analyze a multidimensional surface? No problem! I rollerskate on that shit.Mike Perry2012-06-20
| | | | | | | This commit provides analysis to guide parameter choices for the path bias defense. It contains a standalone attack and defense simulation script, a results.txt with example output, and an observations.txt with comments to help work towards an answer for #6135.
* Merge remote-tracking branch 'origin/master'Mike Perry2012-04-18
|\
| * Make the guard sample rate a consensus param.Mike Perry2011-12-15
| |
| * Be more strict w/ circ failures and fast nodes.Mike Perry2011-12-15
| | | | | | | | | | | | | | | | | | | | | | | | | | We use unfiltered bandwidth because I suspect we're being too generous to fast nodes. To get more feedback, they should have to have spare capacity all the time, not just sometimes. However, gimpy nodes will still get to use filtered bandwidth if we've punished them too much already. Also, circuit failures now target a 0% rate (full success) instead of the network average, but the penalty only applies if you fall below the network average for your node class.
| * Log the network pid error averages.Mike Perry2011-12-15
| | | | | | | | Also, log node class when we cap, and in some other cases.
* | Submodule.Mike Perry2012-04-18
|/
* And it's INT32_MAX, not UINT32_MAX..Mike Perry2011-12-11
| | | | Too bad python doesn't seem to have constants for either?
generated by cgit v1.2.1 (git 2.18.0) at 2024-02-02 10:45:08 +0000