Ray,

You can find where the data is through the following command:

du -h --max-depth=1

Start in the root (/) and follow the trail.

Succes

Best regards,

Joost Waversveld

Ray Leventhal wrote:

Rudi Ahlers wrote:

...

On Wed, Nov 26, 2008 at 4:50 PM, Ray Leventhal centos@swhi.net wrote:

...

Hi all,

Please pardon my newbie-ness on this issue....I've a / partition which is full (quite suddenly, actually) and I'm not sure how to fix this.

I've searched for uneeded logs, etc in /var/log and /tmp to no avail. The system is CentOS 5.2 and is not connected to the internet, serves as a local LAN server running stock stuff...sendmail, dovecot, apache..nothing strange or special going on.

I have additional HDDs available if growing the partition is in order (would appreciate pointers to that, if applicable), but I'm really stumped as to where the space is being eaten up.

Thanks in advance, -Ray

<snip>

Do a search for coredump file (i.e. run updatedb && locate core.*) - they can often fill up the HDD very quicly if something coredumps.

Thanks, Rudi...I hadn't thought of that. The results, however, weren't overly helpful for this issue in this case.

I had about 11M of core dumps from March in the samba/cores directory

The big deal here is that the system is choking...any advice on finding what's making it choke and/or relieving the strain on / would be appreciated. Again, my thanks, -Ray _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos

Sorin Srbu wrote:

Ray Leventhal <> scribbled on Wednesday, November 26, 2008 3:50 PM:

...

I have additional HDDs available if growing the partition is in order (would appreciate pointers to that, if applicable), but I'm really stumped as to where the space is being eaten up.

Try a yum clean all. That might help. But if it's as you say, not connected to the internet, you probably are not running yum at all, so it might not help. Check your temp-directories and clean out as necessary.

You still have 1gig free. Did you do a full install? If you can live with it, uninstall packages you don't need.

---

---

CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos

Hi Sorin,

It was a nearly-full install and I do connect periodically to do updates. I did a yum clean all before writing to the list, with no significant change to the disk usage.

I will list packages next and see what can be removed safely. I truly don't know what is filling the 130G...this should be more than sufficient as I've loaded CentOS5.x on many systems without this issue.

Thanks again for your input, -Ray

Ray Leventhal <> scribbled on Wednesday, November 26, 2008 4:20 PM:

It was a nearly-full install and I do connect periodically to do updates. I did a yum clean all before writing to the list, with no significant change to the disk usage.

I will list packages next and see what can be removed safely. I truly don't know what is filling the 130G...this should be more than sufficient as I've loaded CentOS5.x on many systems without this issue.

Ok, what about the /home-folder, did it reside on a different disk or on the same as where / is? If it's on /, take a look inside /home, your users may have started downloading pictures, mp3 and stuff. You know the drill. ;-)

On Wed, Nov 26, 2008 at 5:50 PM, Ray Leventhal centos@swhi.net wrote:

Sorin Srbu wrote:

...

Ray Leventhal <> scribbled on Wednesday, November 26, 2008 3:50 PM:

...

I have additional HDDs available if growing the partition is in order (would appreciate pointers to that, if applicable), but I'm really stumped as to where the space is being eaten up.

Try a yum clean all. That might help. But if it's as you say, not connected to the internet, you probably are not running yum at all, so it might not help. Check your temp-directories and clean out as necessary.

<snip>

Thanks to all who replied.

/ filled up when my nightly rsync snapshot did something which I'm still looking into.

I run a nightly rsync script to make copies (to an external HDD connected via USB) of user data files:

#backup to USB drive location for /home # /media/bkup is /dev/sdg1 (USB 700GB drive) rsync -av --delete /home/ /media/bkup cd

Well, in /media, there were 2 folders, not just one.../bkup and /bkup_ as well as 2 .lock files. I determined which was the last complete backup and deleted the other... needless to say / space began to increase, but I'm truly puzzled about why a mount point would take up space on / when the media is external.

Anyone with insight into my flawed logic, please let me know :)

Thanks for all help and the ongoing knowledge gained from this list.

-Ray _______________________________________________

That all depends on how the USB HDD was mounted. If it wasn't mounted sucesfully, then /media will be treated as just another folder.

How do you mound the HDD?

Ray Leventhal wrote:

Sorin Srbu wrote:

...

Ray Leventhal <> scribbled on Wednesday, November 26, 2008 3:50 PM:

...

I have additional HDDs available if growing the partition is in order (would appreciate pointers to that, if applicable), but I'm really stumped as to where the space is being eaten up.

Try a yum clean all. That might help. But if it's as you say, not connected to the internet, you probably are not running yum at all, so it might not help. Check your temp-directories and clean out as necessary.

<snip>

Thanks to all who replied.

/ filled up when my nightly rsync snapshot did something which I'm still looking into.

I run a nightly rsync script to make copies (to an external HDD connected via USB) of user data files:

#backup to USB drive location for /home # /media/bkup is /dev/sdg1 (USB 700GB drive) rsync -av --delete /home/ /media/bkup cd

Well, in /media, there were 2 folders, not just one.../bkup and /bkup_ as well as 2 .lock files. I determined which was the last complete backup and deleted the other... needless to say / space began to increase, but I'm truly puzzled about why a mount point would take up space on / when the media is external.

Anyone with insight into my flawed logic, please let me know :)

Thanks for all help and the ongoing knowledge gained from this list.

-Ray _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos

when you mount an device, you will mount it to an existing directory (/media/bkup in this case).

when the drive is mounted, everything that is written to /media/bkup will be written on the external disk.

When the drive is unmounted, everything that is written to that directory, will be written to.. euhm.. the directory, thus the local file system.

So maybe something happened with the mount of the external disk??

Best regards,

Ray Leventhal wrote:

Sorin Srbu wrote:

...

Ray Leventhal <> scribbled on Wednesday, November 26, 2008 3:50 PM:

...

I have additional HDDs available if growing the partition is in order (would appreciate pointers to that, if applicable), but I'm really stumped as to where the space is being eaten up.

Try a yum clean all. That might help. But if it's as you say, not connected to

Hi again all,

There was a 3.5hr power outage last night which explains it all. Sadly, I've got some investigation to do about why my *supposed* 5hrs of battery backup didn't last long enough to cover, but the mount point was, in fact, unmounted and so rsync did it's job right into the folder as opposed to the ext. drive.

As always, my sincere thanks to this list and to our CentOS maintainers.

Kindest regards, -Ray

I noticed the chorus of agreement that your problem was likely a result of failure to mount your backup drive. My backup script, which also uses rsync begins like this, insuring a good mount before shoving bytes around. It's not SUPPOSED to be mounted before the script runs but I test for that, too. (The drive label is OTOT):

#!/bin/bash # Backup using rsync and rotating directories # #Set the Dest. Mount UD=/media/OTOT MS=8388608 # Minimum free space 8GB # # Drive mount logic: # if [ -z $(mount | grep $UD | awk '{ print $3 }') ]; then mount $UD fi if [ -z $(mount | grep $UD | awk '{ print $3 }') ]; then echo "Drive refuses to mount!!" exit 1 fi # # Drive is mounted. Get device name # XDR=$(mount | grep $UD | awk '{ print $1 }')

etc., etc.

On Wed, 2008-11-26 at 11:15 -0500, Ray Leventhal wrote:

Sorin Srbu wrote:

...

<snip>

Hi again all,

There was a 3.5hr power outage last night which explains it all. Sadly, I've got some investigation to do about why my *supposed* 5hrs of battery backup didn't last long enough to cover, but the mount point was, in fact, unmounted and so rsync did it's job right into the folder as opposed to the ext. drive.

My rsync backup script has a check in it to suppress the backup and issue a message if the media is not mounted.

As always, my sincere thanks to this list and to our CentOS maintainers.

Kindest regards, -Ray

<snip sig stuff>

Ray Leventhal wrote:

Sorin Srbu wrote:

...

Ray Leventhal <> scribbled on Wednesday, November 26, 2008 3:50 PM:

...

I have additional HDDs available if growing the partition is in order (would appreciate pointers to that, if applicable), but I'm really stumped as to where the space is being eaten up.

Try a yum clean all. That might help. But if it's as you say, not connected to

Hi again all,

There was a 3.5hr power outage last night which explains it all. Sadly, I've got some investigation to do about why my *supposed* 5hrs of battery backup didn't last long enough to cover, but the mount point was, in fact, unmounted and so rsync did it's job right into the folder as opposed to the ext. drive.

You may want to add some logic to your rsync script to check for a properly mounted disk.