Thread on d.o..

Something to do with the backup / backup migrate module or their configuration.

Thread on d.o.

Something to do with the backup / backup migrate module or their configuration.

It appears the original poster's problem wasn't Drupal being a cancerous tumor but some development version modules being used on a production site with some possible misconfiguration thrown into the mix. (See: http://drupal.org/node/313496)

Maybe someone can let Consumerist know this is unmoderated code submitted by third party contributors. I doubt they'd care. It doesn't make for as good a story. :-/

One thing I can think of: watchdog.

Up until Drupal 5, having a lot of 404s (e.g. from relative images or something) could cause a lot of writes to the watchdog table, and that can add up quickly. But, I can't imagine this causing gigabytes of rows. Moreover, we ship Drupal 5.x with the default setting of keeping only 7 days' worth of watchdog logs.

For Drupal 6.x, we changed that to keep a set number of rows (1000 rows by default).

But the pruning in both cases depend on cron running. So perhaps he did not setup cron correctly? The site would slow to crawl before it uses that much disk space.

Update: This is now reported here. Probably the backup_migrate module, the backup module.

I've contacted Scott Loggins, Justin Jilg, and Gary Kamen at GoDaddy to discuss this matter.

If anyone has a background on the technical reasons for this I'd like to hear it. With hundreds of thousands of Drupal sites world wide and a reported over 5000 Drupal sites at GoDaddy, I am surprised this is the first time I've heard of this.

Kieran

Core dumps (http://en.wikipedia.org/wiki/Core_dump, not drupal core) could do that.
Say there's a bug in ImageMagick, PHP or another binary that is triggered by some requests to drupal, and let's assume this buggy binary starts consuming tons of memory until it finally fails with a core dump, and say this type of behavior is triggered by dozens / hundreds of requests in a short time, then I see how this could have happened.
There are of course many other scenarios. In the end, you'll want to see a detailed analysis of this case. As you say, if you don't have proper limits and safeguards in place, you're just asking for bad news to arrive eventually.

Core dumps (http://en.wikipedia.org/wiki/Core_dump, not drupal core) could do that.
Say there's a bug in ImageMagick, PHP or another binary that is triggered by some requests to drupal, and let's assume this buggy binary starts consuming tons of memory until it finally fails with a core dump, and say this type of behavior is triggered by dozens / hundreds of requests in a short time, then I see how this could have happened.
There are of course many other scenarios. In the end, you'll want to see a detailed analysis of this case. As you say, if you don't have proper limits and safeguards in place, you're just asking for bad news to arrive eventually.

It appears the original poster's problem wasn't Drupal being a cancerous tumor but some development version modules being used on a production site with some possible misconfiguration thrown into the mix. (See: http://drupal.org/node/313496)

Maybe someone can let Consumerist know this is unmoderated code submitted by third party contributors. I doubt they'd care. It doesn't make for as good a story. :-/

http://www.huffingtonpost.com/adam-fendelman/why-i-dont-owe-godaddy-65_b_129276.html

See Why I Don't Owe GoDaddy $6,579.51 and GoDaddy allegedly blames Drupal for running up disk usage and huge hosting bill.

I really find this hard to believe, 250 GB from copying temp files, and this is the first time it's happened? It makes a lot of sense that this is all Adams fault too.

Say there's a bug in ImageMagick, PHP or another binary that is triggered by some requests to drupal, and let's assume this buggy binary starts adtech ile reklam 2.0 dönemi başlıyor ve Trkycmhrytllbtpydrklcktr r10.net seo yarışması consuming tons of memory until it finally fails with a core dump, and say this type of behavior is triggered by dozens / hundreds of requests in a short time, then I see how this could have happened.