Skip to content

Latest commit

 

History

History
65 lines (49 loc) · 3.36 KB

README.md

File metadata and controls

65 lines (49 loc) · 3.36 KB
Raw

ASpace EAD Batch Ingest Scripts

These scripts rely on the presence of the aspace-jsonmodel-from-format plugin in your instance of ArchivesSpace. The plugin converts ead xml to json object model, which allows import into ArchivesSpace. These scripts need not run on the same instance as the aspace instance, we use http post calls to the api (http://your.aspaceinstance:8089/..., for instance)

Prerequisites

  • In your (remote or local) instance of aspace, install the aspace-jsonmodel-from-format plugin. Install info at: https://github.com/lyrasis/aspace-jsonmodel-from-format
  • Full set of EAD files need to be in a local, accessible directory (specified in config.yml, an example is provided here
  • Ruby 2.2+. This should work fine with MRI, jRuby, or whatever, as long as it supports all the dependencies.

Installation

  • Check out this repository

  • Create config.yml file based on config.yml.example

  • Install dependencies via Bundler

    gem install bundler # If not already installed
cd aspace-utils
bundle install

Running the ingester

To run the ingester, place you EAD files in the directory specified in your config.yml, and then run:

bundle exec ingest_aspace.rb

If you want to keep an eye on what it's doing, I recommend:

watch tail ingestlog.log

The ingester populates two log files - ingestlog.log and error_responses

At Harvard, we've been running this under screen to keep this running over long periods of time.

A cautionary note on max_concurrency

This script can do concurrent requests, but versions of ArchivesSpace prior to 1.5.2 had a race condition around creating Subjects/Agents/other shared fields. If using any prior version, you MUST set max_concurrency to 1.

Analysis script

There's also an "analyze_logs.rb" script provided, which can be used thusly:

bundle exec analyze_logs.rb ingestlog.log error_responses > analysis.txt

It currently assumes that there is ONE and only ONE set of logs in each of those files - if you want to use the analysis script, you'll need to wipe ingestlog.log and error_responses between runs.

Note that passing more than two arguments will enter the script in interactive mode - you'll be thrown into a pry session with several interesting local variables defined.

Name Description Type
upload_failures Failures to upload resulting from ASpace DB errors/java errors Hash keyed by approximate cause with values being hashes keyed by eadid
four_hundreds Errors that come from the EAD converter Hash of errors keyed by eadid
five_hundreds Errors from Apache or Java or gremlins Hash of errors keyed by eadid
by_error Big ole hash with 4XX, 5XX, and upload failures aggregated by proximate cause Hash of hashes of arrays, keyed by error class -> proximate cause
ok number of finding aids successfully ingested integer
bad number of finding aids that failed to ingest integer
total number of finding aids processed in total integer

Notes

  • repository ids can be found using the api (http://localhost:8089/repositories, for example); they must be parsed out
  • really this should handle its own log rotation, sorry, PRs welcome or I'll get to it eventually.