Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm installing a program on my Linux server and it stores data locally for a week. However, there is an error that can occur sometimes that will keep it from deleting the data.

To safeguard against this when I installed it on a Unix server I just create a new filesystem of the specified size. However, the filesystem on my linux server was created to have the full partition. Is there some way to set the maximum size for a directory?

Additionally if these was free space in the partition, would it be better to create a new filesystem, or use the above restrictions.

Thanks, Alex

share|improve this question
up vote 2 down vote accepted

I'm not sure of a way to limit the size of a single directory. You could create a new user, assign a quota to them, and then run the process under that user, but I'm guessing that's not what you're after.

As you hint at, you can create a filesystem as a "file" and mount it as the output directory for this app. This would ensure it never spills over to your regular filesystem:

$ dd if=/dev/zero of=~/disk_image_file count=$size_in_blocks

$ mkfs -t ext3 -q ~/disk_image_file

$ mkdir -p ~/mnt/app1/log

$ mount -o loop=/dev/loop0 ~/disk_image_file ~/mnt/app1/log

share|improve this answer
Yeah. I was figuring that if no one at work was suggesting it, it probably didn't exist, but it was worth a shot. Thanks for the code. -Alex – Buzkie Jan 21 '10 at 16:17

You're looking for quotas. Yes, it is entirely possible to implement quotas on most if not all unix filesystems.

Here's what you should read: You shouldn't need to do any kernel config. Any vaguely modern system will likely have this enabled already by the distribution.

Come back if you have problems.

share|improve this answer
Thanks for the suggestion, I was looking into that while waiting for answers. The main problem with quotas is I can't be sure what user the program is running as. In addition, it defaults to run as root, and if I'm deploying to a machine with shared root, I don't want to put a quota on root. – Buzkie Jan 21 '10 at 16:08

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.