11 Nov 2006

Lots and lots of 404s for old content

filed under: journal  :: blogging  :: movable type  :: site news

If you have visited the old pages of this site following links, even links within the site, you would have noticed a ton of 404 (Page not found) errors.

As I've said previously, I did not switch over all the old pages to the new Drupal installation, so this has nothing to do with Drupal.

This is due to one really annoying 'feature' of Movable Type - URLs are not kept in the database. (Since it's not in the database, it doesn't make it to the text backup file either. Incidentally this also applies to Typepad.)

I had to restore my MT installation due to an emergency system reinstall on the server (which meant that the main hard disk had to be reformatted). Restoring the data was fine, since I had MySQL backups. But when I rebuilt the site, I neglected to set the URL string length high enough.

I've fixed that now, but there's another problem. I had several 'custom' URL strings to make them more logical (as you may know, most blogging systems take the individual entry title from the title field of the entry). All of those 'custom' URL strings are now lost.

The only thing I can do is to look at the error logs and manually rename or copy some files to the old filenames. This is a tedious process.

The best way at this point to find old content here is to go to the old content index page and do a text search for whatever you were looking for. Yep, that's a bother too I know.

Drupal 4.7x keeps all 'custom' URL data, or URL aliases, in a table (url_aliases). I'm not sure how other blogging systems handle 'custom' URLs. But if you have a Movable Type installation, tend to use custom URLs a lot, and want real backups with URLs intact, the best thing to do may be to actually back up the static files. Do not assume, as I did, that backing up the database and template files (you do use external template files, don't you?) and then rebuilding if things went wrong would be ok.

Recent popular