Page tree
Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 11 Current »

The Avalon 5.x to 6.x migration tool copies and transforms Avalon 5.x's Fedora 3 data to Avalon 6.x's Fedora 4 data.  The tool can be run as follows:

bundle exec rake avalon:migrate:repo

When you run the migration in the command line, you can see a broad view of how much progress has been made for each Class. Notice the last line indicates a second pass for Migrating MediaObject. The second pass is necessary to associate master files with media objects and get them ordered properly.

Command Options

OptionExplanationExample Usage

Default Value

CONFIRMRequired in order to prevent accidental executionCONFIRM=yes
pidsA comma separated list of Avalon 5.x Fedora 3 pids to migratepids=avalon:1,avalon:2,avalon:3
pidfilePath to a filename containing one Avalon 5.x Fedora 3 pid per linepidfile=/path/to/pidfile

parallel_processesNumber of parallel processes the migration tool should useparallel_processes=4number of processors - 2
overwriteDetermines whether: existing Fedora 4 objects should be kept, cleaned out, and reused (overwrite=true); or new objects should be created even if matching migrated objects already exist (overwrite=false)overwrite=true


namespaceThe Fedora 3 namespace that should be migratednamespace=mediaavalon
skip_completedSkip items with status = completedskip_completed=truefalse

Rerun a migration to bring in new data

You'll want to set skip_completed and overwrite to true. Refer to the logic here

    def migration_required?(pid, klass, method=:migrate)
      status_report = MigrationStatus.find_by(source_class: klass, f3_pid: pid, datastream: nil)
      status_report.nil? ||
        (status_report.status != 'completed' && status_report.status != 'waiting' && method == :migrate) ||
        (status_report.status != 'completed' && method == :second_pass) ||
        (!skip_completed? && overwrite?)

Rerunning the migration

Once the migration is running, it is best to let it fully finish before attempting to rerun failed objects.  The migration should fail objects that are missing parent or child objects.  Once the migration is finished running, a list of the failed object ids can be collected into a file by running this command:

bundle exec rails r 'puts MigrationStatus.where(status: "failed").map(&:f3_pid).uniq.join("\n")' > failed_objects

The migration tool can then be rerun with this list:

bundle exec rake avalon:migrate:repo pidfile=failed_objects

Potential Problems

First read the Migration Report to determine what objects failed migration and why.  

Objects fail with errors about mismatched prefixesTurn off parallel processing by setting parallel_processes=1
Objects fail with LDP conflictTurn off parallel processing by setting parallel_processes=1
Poster and thumbnail images missingSee script on Miscellaneous scripts for R5/R6 migration
Section permalinks incorrectSee script on  Miscellaneous scripts for R5/R6 migration 
Permalink collisionsSee script on  Miscellaneous scripts for R5/R6 migration 

Helper Scripts

See Miscellaneous scripts for R5/R6 migration for scripts that can be adapted to your institution to help find and fix data issues post-migration.

  • No labels