The dry run

A dry run goes through a complete run and changes nothing. It talks to the source site, works out what it would do with every article, and writes that to the log.

During a dry run:

• no article is created, updated or removed;
• no image or file is downloaded;
• no custom field is written.

Use it the first time you set up a task, and every time you change a filter, a mapping or the handling of removed articles.

How to do one

1. Open your task under System > Manage > Scheduled Tasks.
2. Set Dry Run (Preview) to Yes and save.
3. Click Run Test in the task list, or wait for the schedule.

The switch turns itself off when the run has finished. The run after that is a real one. If a dry run stops early, for instance because the source cannot be reached, the switch stays on, so you still get your preview on the next try.

Reading the result

Every line of a dry run starts with [DRY RUN]:

[DRY RUN] No changes will be written to the database.
[DRY RUN] Would download image: /images/news/team.jpg
[DRY RUN] Would create: "Our latest news article" (source: 128)
[DRY RUN] Would update: "Summer opening hours" (source: 97 -> local: 214)
[DRY RUN] Completed - would create: 1 | would update: 1 | would delete: 0 | skipped: 23 | errors: 0

Skipped counts the articles that have not changed since the last run. On a mirror that is up to date, almost everything is skipped. That is the normal picture.

Where the log is

There are two places.

• The task log. Go to System > Manage > Scheduled Tasks and open Execution History. It shows the result of every run, with the summary and every warning and error.
• The log file. With Write Log to File on, every line is also written to administrator/logs/contentsync.php. This is the place for the full detail, line by line.

The Execution History only exists when task logging is switched on in the options of the Scheduled Tasks component.

Log levels

Log Level sets how much is recorded:

• Debug - everything, including every request to the source. Use this while you set a task up or look for a problem.
• Info - what was created, updated, downloaded and removed. The normal setting.
• Warning - only the things that need your attention.
• Error - only failures.

Warnings worth acting on

• A view level, category or author that is not mapped. The article was copied, but with a fallback. Add a mapping or set a default.
• A file that was skipped. It is outside the media folders, too large, or of a type that is not accepted. See Images, documents and media.
• A custom field that is not synced. The field does not exist here, or has another type. See Custom fields.
• Content filtering or SSL verification is switched off. A reminder that you chose a less safe setting.

An error means an article could not be processed. The other articles of that run are still handled. Troubleshooting lists the errors you are most likely to meet.