Skip to content

Latest commit

 

History

History
320 lines (221 loc) · 10 KB

DEPLOYNOTES.rst

File metadata and controls

320 lines (221 loc) · 10 KB
Raw

Deploy and Upgrade notes

1.3

  • Member sorting has been revised, so people should be reindexed in Solr:

    python manage.py index -i person

1.2

  • Revised data exports should be generated:

    python manage.py export_members python manage.py export_events python manage.py export_books

  • To use the new 100 years twitterbot code, you must configure TWITTER_100YEARS in localsettings.py with Twitter API and access credentials for the account where tweets should be posted, and run the twitterbot_100years manage command in schedule mode as a daily cron job every morning before 9am:

    python twitterbot_100years schedule

1.1.1

  • Includes Solr configuration changes to address a stopword search problem on the book search. Copy all files under solr_conf/conf/ to the configured Solr configset and restart Solr to ensure the managed schema is loaded. Then update the schema and reindex:

    python manage.py solr_schema
python manage.py index

1.1

  • Includes updates to Solr managed schema. Copy all files under solr_conf/conf/ to the configured Solr configset and restart Solr to ensure the managed schema is loaded. Then update the schema and reindex:

    python manage.py solr_schema
python manage.py index

  • The migration to calculate purchase dates and adjust start dates for subscription renewals has been revised and should be re-run:

    python manage.py migrate accounts 0032
python manage.py migrate accounts

1.0

XML sitemaps require update to Solr schema to track last modification time of indexed items. Update the schema and reindex everything.

python manage.py solr_schema python manage.py index -c all -i none python manage.py index

0.29

Updates to the book search & browse require a Solr configuration update and work reindex. Copy all files under solr_conf/conf/ to your configured Solr configset and restart Solr to ensure the managed schema is loaded. Then update the schema, clear works from the index, and reindex:

python manage.py solr_schema
python manage.py index -c work -i none
python manage.py index -i work

0.28.2

Django Admin work search requires that the Solr configuration be updated and works reindexed. Copy all files under solr_conf/conf/ to your configured Solr configset and restart Solr to ensure the managed schema is loaded. Then update the schema, clear works from the index, and reindex:

python manage.py solr_schema
python manage.py index -c work -i none
python manage.py index -i work

0.28

The addition of slugs for book urls requires a Solr reindex after running migrations:

python manage.py index -i work

0.27

Bibliographic metadata search for works requires that the Solr configuration be updated and works reindexed. Copy all files under solr_conf/conf/ to your configured Solr configset and restart Solr to ensure the managed schema is loaded. Then update the schema, clear works from the index, and then reindex:

python manage.py solr_schema
python manage.py index -c work -i none
python manage.py index -i work

Data exports for members and events can be generated by manage command:

python manage.py export_members
python manage.py export_events

0.25

  • The name of the canonical Location representing the library should be changed to "Shakespeare and Company" (no &), if it is not already. This change has been reflected in an earlier migration.

0.24

Partial name and accent-agnostic search on the members requires that the Solr configuration be updated and members reindexed. Copy all files under solr_conf/conf/ to your configured Solr configset and restart Solr to ensure the managed schema is loaded. Then update the schema, clear people from the index, and then reindex:

python manage.py solr_schema
python manage.py index -c person -i none
python manage.py index -i person

0.23

  • This release requires library members be cleared and reindexed in order to update the gender field, add the new arrondissement field for use in faceting, and ensure all members have slugs for urls:

    python manage.py index -c person -i person

  • This release adds new Page content types for blog-style essay content and a landing page for blog entries. If content exists that should be migrated to these pages, new pages with the correct content type must be created.

  • You must set MAPBOX_ACCESS_TOKEN, MAPBOX_BASEMAP, and PARIS_OVERLAY in local_settings.py in order to render the address map shown on member detail pages. See the note in local_settings.py.sample for more information.

0.22

  • Member search filter on nationality and the new card browse require a Solr schema update and reindex:

    python manage.py solr_schema python manage.py index -i person python manage.py index -i card

0.21

  • You must configure GTAGS_ANALYTICS_ID and GTAGS_ANALYTICS_ENV in local_settings.py in order for Google Analytics to function. See the note in local_settings.py.sample for more information.

  • You must configure CSP_REPORT_ONLY and CSP_REPORT_URI in local_settings.py in order for Content-Security Policy to function. See the note in local_settings.py.sample for more information and suggested settings for your environment. Note that the policies themselves are defined in the settings.py file but can be overridden in local_settings.py, which you may want to do when developing.

  • After this version is deployed, run import_figgy_cards to migrate old pudl image urls in Bibliography and Footnote records to use the new objects in PUL's Figgy. The CSV file mapping old paths to new Figgy ids is included as a fixture for convenience:

    python manage.py import_figgy_cards mep/accounts/fixtures/pudl-to-figgy-mapping.csv

0.17

  • This release requires a reindex to update the birth and death year fields for use in faceting.

0.16

  • This release requires a reindex to incorporate a bugfix for account partial date handling.

0.15

  • This release requires a reindex to update the sex field for use in faceting.

0.14

  • You must configure OCLC_WSKEY in localsettings.py before you can use the new reconcile_oclc manage command. You should also configure a TECHNICAL_CONTACT email address.

0.13

  • Member search requires an updated solrconfig.xml. Copy solr_conf/conf/solrconfig.xml to your configured Solr configset directory.

0.12

  • Index book data in Solr to populate the book browse:

    python manage.py index -i item

0.11

  • Solr connection information should be configured in local settings as a SOLR_CONNECTION. See the sample Solr config in local_settings.py.sample.

  • The Solr configuration in the solr_conf directory must be installed on the Solr server the configsets directory prior to deployment with a name matching the CONFIGSET in the SOLR_CONNECTIONS default configuration. See Solr Config Sets documentation for more details. Possible sequence of commands, starting from the server where the mep-django codebase is checked out:

    scp -r mep-django/solr_config solr-server:sandco

    Then on the server where Solr is running:

    mkdir -p /var/lib/solr/data/configsets
cp -r sandco /var/lib/solr/data/configsets/
chown solr:solr -R /var/lib/solr/data/configsets

  • After the configset is in place on the server, run solr_schema to configure the configured Solr core. This command will create the core with the configured CONFIGSET if the core does not already exist:

    python manage.py solr_schema

  • Index member data into Solr:

    python manage.py index

0.10

  • Switching from Mezzanine to Wagtail requires a manual migration before installing the new version to avoid migration dependency conflicts:

    python manage.py migrate pages zero

  • Wagtail provides predefined groups for Editor and Moderator. Users who were previously in the Content Editor group should be added to one of these, and the Content Editor group should be removed.

  • Run python manage.py setup_site_pages to create stub pages for all site content needed for main site navigation.

0.6 Borrowing events and Title stubs

0.3 Data Entry Improvements

  • This update adds Mezzanine and enables the Django Site framework. The default Site should be configured after migrations complete.

0.2 Logbook admin

  • Logbook data should be imported from XML using the

import_logbooks manage command. The logbook XML is located in https://github.com/Princeton-CDH/mapping-expatriate-paris under transcriptions/logbooks/*.xml. The command tasks a list of files using standard shell globbing. The import should be from the develop branch, which has been cleaned up for import.

0.1 Personography admin

  • The application must be configured with a GeoNames username and Mapbox access token in order to use GeoNames and Mapbox APIs.
  • Personography data should be imported from XML using the import_personography manage command. The personography XML for import is included in https://github.com/Princeton-CDH/mapping-expatriate-paris under transcriptions/personography.xml NOTE: import should be run from the develop branch, which has the latest version and has been cleaned up for import.