Release 1.6.1

This is a patchlevel release with the following fixes:

• Skip connection-drain wait for development BigCouch
• Skip Ingestion 2 operations with legacy inventory
• Fix variable inclusion, Omeka deployment

Release 1.6.0

• Enable CORS for API redirects

Release 1.5.0

• Assign kernel overcommit setting on workers
• Have Marmotta's NGINX log request time
• Lower the number of PostgreSQL backups kept

Release 1.4.0

• Heidrun: add base URI for providers to settings
• Disable Marmotta's LDCache module
• Set up custom log rotation for Marmotta's Nginx
• Use file-based caching for API in contentqa
• Improve Tomcat logfile rotation

Release 1.3.0

• Change solr configuration in ingestion_app role
• Add param for controlling Munin alert sending
• Install php5-cli for common_php role [fixes #7788]
• Set Linux kernel params for memory overcommitment
• Add an elasticsearch section to Heidrun's settings.local.yml

Release 1.2.0

• Add standalone ingestion app deployment playbooks
• Fix deployment of heidrun Gemfile in development
• Fix incorrect rsync options for deployments
• Support new vendored deployment for Heiðrún mappings
• Improve cooperation between local and git deployments
• Other minor fixes

Upgrading

You may, optionally, add the new heidrun_mappings_branch_or_tag variable to your custom files in ansible/roles/ingestion_app/vars

Release 1.1.0

• Add Ingestion 2, our new ingestion system
• Make PostgreSQL maintenance playbooks more flexible
• Deploy 'master' branch of Exhibitions by default
• Add Travis CI configuration
• Fix: remove old Cloudant package list for BigCouch
• Fix redundant reboot of PostgreSQL server

Release 1.0.0

• Rewrite sitemap URL on site proxies
• Add rotation of various logfiles and improve logging
• Add PHP-FPM tuning parameters
• Improve Unicorn init scripts and restarts
• Improve NGINX tuning
• Fix Omeka deployment
• Add relational database dumps
• Improve upgrades and reboots
• Improve frontend deployments
• Improve CMS and PostgreSQL restarts
• Require Ansible 1.8, primarily for connection draining support
  • Requires upgrade to Ansible 1.8 for existing installations.
• Improve specification of PostgreSQL database encoding
  • Requires update of all variables file for existing installations.
• Introduce S3 (cloud storage) for frontend
  • Requires Amazon S3 account to run frontend app
• Fix BigCouch installation to circumvent removal of original package repository files.
• Various other fixes

Upgrading:

• Upgrade Ansible to version 1.8
• Set up an Amazon S3 account if you want to use the frontend application and update group_vars/all with the aws_* variables given in the all.dist file.
• Update your group_vars/all file with a value for dpla_locale. (See all.dist.)

Release 0.3.0

• Fix munin-node access control
• Add Munin charts for memcached
• Add template/install task for is_ingestion_running.sh script
• Add BigCouch/ES maintenance scripts on central (refs #7738)
• Add rotation of API and frontend logs (fixes #7737)
• Fix: add 'logrotate' tag to include statements
• Improve robots.txt
• Redirect "www" requests
• Add notifications to Munin and Slack notifier script
• Add site proxy cache clearing to API/frontend deployment
• Use /public as frontend app root

Release 0.2.0

• Add vagrant-vbguest prerequisite
• Add bigdesk plugin to Elasticsearch
• Add playbook for central server (not relevant for development)
• Add loadbalancer management for staging, production
• Rearrange and add playbooks for greater convenience
• Have contentqa_all.yml include new common playbook
• Add playbook to initialize new development environment
• Break out common httpd into separate playbook
• Add ntp to common packages
• Update required Ansible version
• Add munin monitoring
• Improve frontend and api deployments
• Add deployment playbooks for CMS roles
• Add playbooks for clearing caches
• Add requirements.txt for non-Ansible Python dependencies
• Add reboot and upgrade playbooks (not for development)
• Add max shared memory setting for PostgreSQL
• Plus various minor (non-development) additions and various bugfixes.

Upgrading:

• Upgrade Ansible to at version 1.7 if it is not there already (see README.md)
• Make sure Vagrant is at version 1.6 (see README.md)
• Make sure Python packages are up-to-date, using pip as documented in README.md
• Install vagrant-vbguest Vagrant plugin if it is not already installed (see README.md)