Start a Conversation

Unsolved

This post is more than 5 years old

5023

September 4th, 2013 13:00

Celerra /nbsnas/var/dump directory is full.

Hello all,

I have an older Celerra running 6.0.43-0 that is showing the /nbs/nas/var/dump directory is 100% full. When looking at the directory the lost+found file seems to be the issue. I have a few screen shots below.

Does anyone have any ideas on how to fix this issue? I am not really sure what the lost+found file is actually for in this directory. Any advice is appreciated.

9-4-2013 4-44-38 PM Celerra Dump File Full.png

9-4-2013 4-48-34 PM Celerra directory full.png

2 Intern

 • 

20.4K Posts

September 4th, 2013 14:00

what else is in that directory ? i find this directory contains kernel dump files if datamover happens to panic which can be safe to delete after they have been analyzed by EMC and copied over.

2 Posts

September 5th, 2013 08:00

I am not able to list all the files in that directory. It shows that there is a total of 20, but only 2 show up.

2 Intern

 • 

20.4K Posts

September 5th, 2013 11:00

can you su to root account and "ls -la" again ?

su root  (not su - root)

86 Posts

September 9th, 2014 13:00

dynamox, I'm having very similar issue on my NS-120. When I issue the "ls -la" using root, I see 2 directories, "lost+found" & "mgmtd".

I did have a bunch of old EMC Support log files that I deleted. I also deleted thousands of old alerts, but when I do "df -h" it still shows that my /nbsnas/var/dump" directory is 100% utilized. Any other thoughts?

2 Intern

 • 

20.4K Posts

September 9th, 2014 14:00

check inside of /nbsnas/var/backup

this folder contains nas database backup files, there should be no more than a day or two worth of tar.gz files. If you see more talk to support, they should get cleaned up automatically.

Also check /nbsnas/var/emcsupport  , if you see a bunch of old support_materials files, delete them.

8.6K Posts

September 9th, 2014 14:00

Talk to EMC support – they have a list / knowledgebase article on what to clean up

86 Posts

September 9th, 2014 15:00

Thanks for the quick response! When I go to that directory  do a "ls -ltr", it says there are 38460 files/directories. It's displaying 18 .gz files though. Think I should call support?

I'm not a unix guy, so I'm a little confused why I'm only seeing 18 files, but it says "38460" total.

2 Intern

 • 

20.4K Posts

September 9th, 2014 18:00

total is how many 1k blocks are in that directory (not recursive), to see more humanly readable output run  ls -rtlh

so looks like that directory is fine if you only see 18 files. Did you check emcsupport directory ?

2 Intern

 • 

20.4K Posts

September 10th, 2014 06:00

cd into /nbsnas/var/dump and run du -h

this should give break down by each subfolder, do you see which one is taking up a lot of space ?

2 Intern

 • 

20.4K Posts

September 10th, 2014 06:00

how's total utilization now ?

86 Posts

September 10th, 2014 06:00

/dev/mapper/emc_vg_lun_0-emc_lv_nas_var_dump

                      1.7G  1.7G     0 100% /nbsnas/var/dump

2 Intern

 • 

20.4K Posts

September 10th, 2014 06:00

and from the command line ?

df -h | grep dump

86 Posts

September 10th, 2014 06:00

The same. I had deleted those emcsupport files before posting in this thread so nothing else has changed. Within Unisphere, I'm still seeing an alert every hour that states the following:

dskMon[7541]: FS /dev/mapper/emc_vg_lun_0-emc_lv_nas_var_dump mounted on /nbsnas/var/dump filling up (at 100%, max = 90%).

86 Posts

September 10th, 2014 06:00

I did delete a bunch of .zip files in the emcsupport directory. Using "ls -rtlh", I'm seeing total 15M & 1 .zip file.

86 Posts

September 10th, 2014 07:00

Output:

16K     ./lost+found

1.6G    ./mgmtd

1.6G    .

Looks like there is a 1.6GB file in mgmt called "core.13555"

No Events found!

Top