Check could really use some love. Any volunteers please?
--Roger

https://lists.torproject.org/pipermail/tor-talk/2013-August/029306.html

See /docs for an idea of what's going on here.

Generating the exit list requires stem, Tor's python controller library. Assuming you have a virtualenv ready, just:

pip install -r requirements.txt

Alternatively,

apt-get install python-dateutil python-stem  # stem 1.1 is in unstable

For the server itself, you'll need go and gettext. Installing that might look like:

apt-get install git golang gettext
go get github.com/samuel/go-gettext/gettext

Then you can run make and wait for git and rsync to fetch all the data and launch the server.

Please run the tests before sending a pull request:

make test

The data that make start pulls in will quickly become stale. What you want to do is run a tor instance with the following configurations in your torrc:

FetchDirInfoEarly 1
FetchDirInfoExtraEarly 1
FetchUselessDescriptors 1
UseMicrodescriptors 0
DownloadExtraInfo 1

Then setup a cron job to run a script like scripts/cpexits.sh every hour. Setting up TorDNSEL to get the exit addresses is beyond the scope of this readme.

Assuming debian, install the dependencies,

apt-get install git golang gettext python-dateutil python-stem
go get github.com/samuel/go-gettext/gettext

The cron job and init script assume a base directory of /opt/check.

git clone https://git.torproject.org/check.git /opt/check
cd /opt/check
make i18n
make exits
make install

Then you start it up,

/etc/init.d/check start

The production check.tpo symlinks TorDNSEL's state file, exit-addresses, to its DocumentRoot. This is to replace what was formerly at exitlist.tpo

Translations are maintained in Transifex. See Tor's translation overview to get involved.

54.7 requests/sec - 47.0 kB/second - 879 B/request
487 requests currently being processed, 63 idle workers
that's check right now (43 million hits over the past 9 days)

MIT