This rewrites the top-level loop of hydra-evaluator in C++. The Perl
stuff is moved into hydra-eval-jobset. (Rewriting the entire evaluator
would be nice but is a bit too much work.) The new version has some
advantages:
* It can run multiple jobset evaluations in parallel.
* It uses PostgreSQL notifications so it doesn't have to poll the
database. So if a jobset is triggered via the web interface or from
a GitHub / Bitbucket webhook, evaluation of the jobset will start
almost instantaneously (assuming the evaluator is not at its
concurrency limit).
* It imposes a timeout on evaluations. So if e.g. hydra-eval-jobset
hangs connecting to a Mercurial server, it will eventually be
killed.
makeBinPath takes care to use the correct output.
nix-repl> lib.makeSearchPath "bin" [pkgs.nix]
"/nix/store/zpp83pr21ihxwsr15l6mkzwkr49zj71d-nix-1.11.2-dev/bin"
nix-repl> lib.makeBinPath [pkgs.nix]
"/nix/store/9n8c3g541qn43yjjs94f1a0m69wp8scg-nix-1.11.2/bin"
The required configuration in hydra.conf:
enable_google_login = 1
google_client_id = 238429sdjkds....apps.googleusercontent.com
and optionally persona_allowed_domains to restrict to one or more
domains.
This is currently done by a separate program that periodically
calls "hydra-queue-runner --status". Eventually, I'll do this
in the queue runner directly.
Fixes#220.
Fixes errors like:
Caught exception in engine "Wide character in syswrite at /nix/store/498lwsrn5kkdh1q8kn3vcpd3457w6m7a-hydra-perl-deps/lib/perl5/site_perl/5.16.3/Starman/Server.pm line 547."
Note that these errors didn't happen if the database encoding was set
to SQL_ASCII (which was the case for hydra.nixos.org, explaining why
it didn't get these errors). However, now the encoding must be
UTF8. To change it, do:
update pg_database set encoding = pg_char_to_encoding('UTF8') where datname = 'hydra';
