move ansible directory to top level; generalize formerly vagrant-specific ansible configuration; let upstart manage logging with "console log"

ansible/hosts-vagrant

@@ -0,0 +1,23 @@
+ansible_ssh_private_key_file=.vagrant/machines/10.9.9.9/virtualbox/private_key
+warcs_dir=/vagrant/warcs
+brozzler_pip_name='-e git+file:///brozzler'
+
+### possible values for a prod deployment
+# brozzler_pip_name=brozzler  # get it from pypi
+# brozzler_pip_name=git+https://github.com/internetarchive/brozzler.git#egg=brozzler
+
+[rethinkdb]
+10.9.9.9
+
+[warcprox]
+work_dir=/vagrant
+10.9.9.9
+
+[brozzler-worker]
+10.9.9.9
+
+[brozzler-webconsole]
+10.9.9.9
+
+[pywb]
+10.9.9.9

ansible/playbook.yml

@@ -1,9 +1,4 @@
 ---
-- name: apply common configuration to all nodes
-  hosts: all
-  roles:
-  - common
-
 - name: deploy rethinkdb
   hosts: rethinkdb
   roles:

ansible/roles/brozzler-webconsole/meta/main.yml

@@ -0,0 +1,3 @@
+---
+dependencies:
+- role: common

ansible/roles/brozzler-webconsole/tasks/main.yml

@@ -1,9 +1,15 @@
 ---
+- name: mkdir {{venv_root}}/brozzler-webconsole-ve34
+  file: path={{venv_root}}/brozzler-webconsole-ve34 state=directory
+        owner={{user}}
+  become: true
 - name: install brozzler[webconsole] in virtualenv
-  pip: name='-e /brozzler[webconsole]'
-       virtualenv=/home/vagrant/brozzler-webconsole-ve34
+  pip: name='{{brozzler_pip_name}}[webconsole]'
+       virtualenv={{venv_root}}/brozzler-webconsole-ve34
        virtualenv_python=python3.4
        extra_args='--no-input --upgrade --pre'
+  become: true
+  become_user: '{{user}}'
   notify:
   - restart brozzler-webconsole
 - name: install upstart config /etc/init/brozzler-webconsole.conf

ansible/roles/brozzler-webconsole/templates/brozzler-webconsole.conf.j2

@@ -3,16 +3,16 @@ description "brozzler-webconsole"
 start on runlevel [2345]
 stop on runlevel [!2345]
 
-env PYTHONPATH=/home/vagrant/brozzler-webconsole-ve34/lib/python3.4/site-packages
-env PATH=/home/vagrant/brozzler-webconsole-ve34/bin:/usr/bin:/bin
+env PYTHONPATH={{venv_root}}/brozzler-webconsole-ve34/lib/python3.4/site-packages
+env PATH={{venv_root}}/brozzler-webconsole-ve34/bin:/usr/bin:/bin
 env LC_ALL=C.UTF-8
 
 env WAYBACK_BASEURL=http://{{groups['pywb'][0]}}:8880/brozzler
 env RETHINKDB_SERVERS={{groups['rethinkdb'] | join(',')}}
 env RETHINKDB_DB=brozzler
 
-setuid vagrant
+setuid {{user}}
 
-# console log
+console log
 
-exec gunicorn --bind=0.0.0.0:8881 brozzler.webconsole:app >>/vagrant/logs/brozzler-webconsole.log 2>&1
+exec gunicorn --bind=0.0.0.0:8881 brozzler.webconsole:app

ansible/roles/brozzler-worker/meta/main.yml

@@ -0,0 +1,3 @@
+---
+dependencies:
+- role: common

ansible/roles/brozzler-worker/tasks/main.yml

@@ -37,22 +37,32 @@
   become: true
   notify:
   - restart Xvnc
+- name: mkdir {{venv_root}}/websockify-ve34
+  become: true
+  file: path={{venv_root}}/websockify-ve34 state=directory owner={{user}}
 - name: install websockify in virtualenv
   pip: name=git+https://github.com/kanaka/websockify.git#egg=websockify
-       virtualenv=/home/vagrant/websockify-ve34
+       virtualenv={{venv_root}}/websockify-ve34
        virtualenv_python=python3.4
        extra_args='--no-input --upgrade --pre'
+  become: true
+  become_user: '{{user}}'
 - name: install vnc-websock upstart config /etc/init/vnc-websock.conf
   template: src=templates/vnc-websock.conf.j2 dest=/etc/init/vnc-websock.conf
   become: true
   notify:
   - restart vnc-websock
+- name: mkdir {{venv_root}}/brozzler-ve34
+  become: true
+  file: path={{venv_root}}/brozzler-ve34 state=directory owner={{user}}
 - name: install brozzler in virtualenv
   pip: # name=git+https://github.com/internetarchive/brozzler.git#egg=brozzler
-       name='-e /brozzler'
-       virtualenv=/home/vagrant/brozzler-ve34
+       name='{{brozzler_pip_name}}'
+       virtualenv={{venv_root}}/brozzler-ve34
        virtualenv_python=python3.4
        extra_args='--no-input --upgrade --pre'
+  become: true
+  become_user: '{{user}}'
   notify:
   - restart brozzler-worker
 - name: install brozzler-worker upstart config /etc/init/brozzler-worker.conf

ansible/roles/brozzler-worker/templates/Xvnc.conf.j2

@@ -3,11 +3,11 @@ description "Xvnc"
 start on runlevel [2345]
 stop on runlevel [!2345]
 
-setuid vagrant
+setuid {{user}}
 
 console log
 
-exec nice Xvnc4 :1 -auth /tmp/Xauthority.vagrant \
+exec nice Xvnc4 :1 -auth /tmp/Xauthority.{{user}} \
     -geometry 1600x1000 -depth 24 -rfbwait 0 -nolisten tcp -rfbport 5901 \
     -SecurityTypes None -pn -fp /usr/share/fonts/X11/misc/ -co /etc/X11/rgb \
     AcceptCutText=0 AcceptPointerEvents=0 AcceptKeyEvents=0

ansible/roles/brozzler-worker/templates/brozzler-worker.conf.j2

@@ -4,13 +4,13 @@ start on runlevel [2345]
 stop on runlevel [!2345]
 
 env DISPLAY=:1
-env PATH=/home/vagrant/brozzler-ve34/bin:/usr/bin:/bin
-env PYTHONPATH=/home/vagrant/brozzler-ve34/lib/python3.4/site-packages
+env PATH={{venv_root}}/brozzler-ve34/bin:/usr/bin:/bin
+env PYTHONPATH={{venv_root}}/brozzler-ve34/lib/python3.4/site-packages
 env LANG=C.UTF-8
 
-setuid vagrant
+setuid {{user}}
 
-# console log
+console log
 
 # depends on vnc server
 start on started Xvnc
@@ -20,4 +20,4 @@ kill timeout 60
 
 exec nice brozzler-worker \
     --rethinkdb-servers={{groups['rethinkdb'] | join(',')}} \
-    --max-browsers=4 >>/vagrant/logs/brozzler-worker.log 2>&1
+    --max-browsers=4

ansible/roles/brozzler-worker/templates/vnc-websock.conf.j2

@@ -3,12 +3,12 @@ description "vnc-websock"
 start on runlevel [2345]
 stop on runlevel [!2345]
 
-setuid vagrant
+setuid {{user}}
 
 console log
 
-env PYTHONPATH=/home/vagrant/websockify-ve34/lib/python3.4/site-packages
-env PATH=/home/vagrant/websockify-ve34/bin:/usr/bin:/bin
+env PYTHONPATH={{venv_root}}/websockify-ve34/lib/python3.4/site-packages
+env PATH={{venv_root}}/websockify-ve34/bin:/usr/bin:/bin
 
 # port 8901 is hard-coded in brozzler/webconsole/static/partials/workers.html
 exec nice websockify 0.0.0.0:8901 localhost:5901

ansible/roles/common/defaults/main.yml

@@ -0,0 +1,7 @@
+# variables default values, these can be overridden in the ansible inventory
+# or various other places
+---
+user: brozzler
+venv_root: /opt
+warcs_dir: /var/tmp/warcs
+brozzler_pip_name: brozzler   # get it from pypi by default

ansible/roles/common/tasks/main.yml

@@ -19,6 +19,7 @@
 ##   command: python3 setup.py install chdir=/tmp/pip-8.1.2
 ##            creates=/usr/local/lib/python2.7/dist-packages/pip-8.1.2-py2.7.egg/pip/__init__.py
 ##   become: true
-- name: mkdir /vagrant/logs
-  file: path=/vagrant/logs state=directory
+- name: ensure service user {{user}} exists
+  user: name={{user}} system=yes createhome=no home=/nonexistent
+        shell=/usr/sbin/nologin
   become: true

ansible/roles/pywb/meta/main.yml

@@ -0,0 +1,3 @@
+---
+dependencies:
+- role: common

ansible/roles/pywb/tasks/main.yml

@@ -1,16 +1,24 @@
 ---
+- name: mkdir {{venv_root}}/pywb-ve34
+  file: path={{venv_root}}/pywb-ve34 state=directory
+        owner={{user}}
+  become: true
 - name: install pywb in virtualenv
   pip: name=pywb
-       virtualenv=/home/vagrant/pywb-ve34
+       virtualenv={{venv_root}}/pywb-ve34
        virtualenv_python=python3.4
        extra_args='--no-input --upgrade --pre'
+  become: true
+  become_user: '{{user}}'
   notify:
   - restart pywb
 - name: install brozzler in pywb virtualenv
-  pip: name='-e /brozzler'
-       virtualenv=/home/vagrant/pywb-ve34
+  pip: name='{{brozzler_pip_name}}'
+       virtualenv={{venv_root}}/pywb-ve34
        virtualenv_python=python3.4
        extra_args='--no-input --upgrade --pre'
+  become: true
+  become_user: '{{user}}'
   notify:
   - restart pywb
 - name: pywb config file /etc/pywb.yml

ansible/roles/pywb/templates/pywb.conf.j2

@@ -0,0 +1,14 @@
+description "pywb"
+
+start on runlevel [2345]
+stop on runlevel [!2345]
+
+env PYTHONPATH={{venv_root}}/pywb-ve34/lib/python3.4/site-packages
+env PATH={{venv_root}}/pywb-ve34/bin:/usr/bin:/bin
+env PYWB_CONFIG_FILE=/etc/pywb.yml
+
+setuid {{user}}
+
+console log
+
+exec nice brozzler-wayback

ansible/roles/pywb/templates/pywb.yml.j2

@@ -1,9 +1,12 @@
-archive_paths: /vagrant/warcs/
+archive_paths: {{warcs_dir}}
 collections:
   brozzler:
     index_paths: !!python/object:brozzler.pywb.RethinkCDXSource
       db: brozzler
-      servers: [localhost]
+      servers:
+      {% for node in groups['rethinkdb'] %}
+      - {{node}}
+      {% endfor %}
       table: captures
 enable_auto_colls: false
 enable_cdx_api: true

ansible/roles/rethinkdb/tasks/main.yml

@@ -11,13 +11,17 @@
   become: true
   notify:
   - restart rethinkdb
-# XXX rethinkdb fails to start in spite of this, I think because /vagrant
-# gets mounted too late, and it tries to log there
 - name: ensure rethinkdb starts on reboot
   service: name=rethinkdb enabled=yes
+- stat: path=/var/log/rethinkdb.log
+  register: p
+- name: ensure user rethinkdb owns /var/log/rethinkdb.log
+  file: path=/var/log/rethinkdb.log owner=rethinkdb state=touch mode=0644
+  when: not p.stat.exists
+  become: true
 - name: ensure rethinkdb instance config file is installed
-  template: src=templates/rethinkdb-brozzler-vagrant-1.conf.j2
-            dest=/etc/rethinkdb/instances.d/rethinkdb-brozzler-vagrant-1.conf
+  template: src=templates/rethinkdb-brozzler.conf.j2
+            dest=/etc/rethinkdb/instances.d/rethinkdb-brozzler.conf
   become: true
   notify:
   - restart rethinkdb

ansible/roles/rethinkdb/templates/rethinkdb-brozzler.conf.j2

@@ -0,0 +1,6 @@
+bind=0.0.0.0
+# directory=/var/lib/rethinkdb
+log-file=/var/log/rethinkdb.log
+{% for node in groups['rethinkdb'] %}
+join={{node}}:29015
+{% endfor %}

ansible/roles/warcprox/defaults/main.yml

@@ -0,0 +1,2 @@
+---
+work_dir: /var/tmp

ansible/roles/warcprox/handlers/main.yml

@@ -0,0 +1,4 @@
+---
+- name: restart warcprox
+  service: name=warcprox state=restarted
+  become: true

ansible/roles/warcprox/meta/main.yml

@@ -0,0 +1,3 @@
+---
+dependencies:
+- role: common

ansible/roles/warcprox/tasks/main.yml

@@ -11,11 +11,16 @@
   - libssl-dev
   - tor
   - git
+- name: mkdir {{venv_root}}/warcprox-ve34
+  become: true
+  file: path={{venv_root}}/warcprox-ve34 state=directory owner={{user}}
 - name: install warcprox in virtualenv
   pip: name=git+https://github.com/internetarchive/warcprox.git@2.x#egg=warcprox
-       virtualenv=/home/vagrant/warcprox-ve34
+       virtualenv={{venv_root}}/warcprox-ve34
        virtualenv_python=python3.4
        extra_args='--no-input --upgrade --pre'
+  become: true
+  become_user: '{{user}}'
   notify:
   - restart warcprox
 - name: install upstart config /etc/init/warcprox.conf

ansible/roles/warcprox/templates/warcprox.conf.j2

@@ -0,0 +1,24 @@
+description "warcprox"
+
+start on runlevel [2345]
+stop on runlevel [!2345]
+
+env PYTHONPATH={{venv_root}}/warcprox-ve34/lib/python3.4/site-packages
+env PATH={{venv_root}}/warcprox-ve34/bin:/usr/bin:/bin
+
+# by default warcprox creates some files/dirs relative to cwd
+chdir {{work_dir}}
+setuid {{user}}
+
+console log
+
+# --profile
+exec nice warcprox \
+         --dir={{warcs_dir}} \
+         --base32 \
+         --gzip \
+         --rollover-idle-time=180 \
+         --onion-tor-socks-proxy=localhost:9050 \
+         --rethinkdb-servers={{groups['rethinkdb']|join(',')}} \
+         --rethinkdb-db=brozzler \
+         --rethinkdb-big-table

setup.py

@@ -32,7 +32,7 @@ def find_package_data(package):
 
 setuptools.setup(
         name='brozzler',
-        version='1.1b7.dev96',
+        version='1.1b7.dev97',
         description='Distributed web crawling with browsers',
         url='https://github.com/internetarchive/brozzler',
         author='Noah Levitt',

vagrant/Vagrantfile

@@ -7,7 +7,7 @@ Vagrant.configure(2) do |config|
   config.vm.synced_folder "..", "/brozzler"
 
   config.vm.provision "ansible" do |ansible|
-    ansible.inventory_path = "ansible/hosts"
-    ansible.playbook = "ansible/playbook.yml"
+    ansible.inventory_path = "../ansible/hosts-vagrant"
+    ansible.playbook = "../ansible/playbook.yml"
   end
 end

vagrant/ansible/hosts

@@ -1,16 +0,0 @@
-ansible_ssh_private_key_file=.vagrant/machines/10.9.9.9/virtualbox/private_key
-
-[rethinkdb]
-10.9.9.9
-
-[warcprox]
-10.9.9.9
-
-[brozzler-worker]
-10.9.9.9
-
-[brozzler-webconsole]
-10.9.9.9
-
-[pywb]
-10.9.9.9

vagrant/ansible/roles/pywb/templates/pywb.conf.j2

@@ -1,14 +0,0 @@
-description "pywb"
-
-start on runlevel [2345]
-stop on runlevel [!2345]
-
-env PYTHONPATH=/home/vagrant/pywb-ve34/lib/python3.4/site-packages
-env PATH=/home/vagrant/pywb-ve34/bin:/usr/bin:/bin
-env PYWB_CONFIG_FILE=/etc/pywb.yml
-
-setuid vagrant
-
-# console log
-
-exec nice brozzler-wayback >>/vagrant/logs/pywb.log 2>&1

vagrant/ansible/roles/rethinkdb/templates/rethinkdb-brozzler-vagrant-1.conf.j2

@@ -1,5 +0,0 @@
-runuser=vagrant
-bind=0.0.0.0
-# directory=/var/lib/rethinkdb
-# log-file=/var/log/rethinkdb.log
-log-file=/vagrant/logs/rethinkdb.log  # synced dir

vagrant/ansible/roles/warcprox/handlers/main.yml

@@ -1,14 +0,0 @@
----
-# - name: start warcprox
-#   environment:
-#     PYTHONPATH: /home/vagrant/warcprox-ve34/lib/python3.4/site-packages
-#     PATH: /home/vagrant/warcprox-ve34/bin:/usr/bin:/bin
-#   args:
-#     executable: /bin/bash
-#   shell: nice warcprox --dir=/vagrant/warcs --base32 --gzip
-#          --rollover-idle-time=180 --cacert=/vagrant/warcprox-ca.pem
-#          --onion-tor-socks-proxy=localhost:9050 --rethinkdb-servers=localhost
-#          --rethinkdb-big-table &> /vagrant/logs/warcprox.out &
-- name: restart warcprox
-  service: name=warcprox state=restarted
-  become: true

vagrant/ansible/roles/warcprox/templates/warcprox.conf.j2

@@ -1,26 +0,0 @@
-description "warcprox"
-
-start on runlevel [2345]
-stop on runlevel [!2345]
-
-env PYTHONPATH=/home/vagrant/warcprox-ve34/lib/python3.4/site-packages
-env PATH=/home/vagrant/warcprox-ve34/bin:/usr/bin:/bin
-
-# by default warcprox creates some files/dirs relative to cwd
-chdir /home/vagrant
-setuid vagrant
-
-# console log
-
-# --profile
-exec nice warcprox \
-         --dir=/vagrant/warcs \
-         --base32 \
-         --gzip \
-         --rollover-idle-time=180 \
-         --cacert=/vagrant/warcprox-ca.pem \
-         --onion-tor-socks-proxy=localhost:9050 \
-         --rethinkdb-servers=localhost \
-         --rethinkdb-db=brozzler \
-         --rethinkdb-big-table >>/vagrant/logs/warcprox.log 2>&1
-         # --rethinkdb-servers={{groups['rethinkdb'] | join(',')}} \