Unsolved

This post is more than 5 years old

1 Rookie

 • 

13 Posts

2598

November 11th, 2008 13:00

Accidently deleted .etc and lost+found from file system

Hi everyone!

While running a robocopy job from a Windows file server to a new file system on our NS20, I forgot to tell it to ignore the .etc and lost+found directories that are on the root of a CIFS shared file system. I guess I have two questions: Is this a problem and if so, how do I fix it. I'm still new to the Celerra product so go easy on me!

11 Legend

 • 

20.4K Posts

 • 

87.4K Points

November 11th, 2008 13:00

i love option two in that solution ..so casually ..just reboot your datamover.

108 Posts

November 11th, 2008 13:00

Hey,

Take a look at emc66658 :-)

Moderator

 • 

285 Posts

November 11th, 2008 14:00

i love option two in that solution ..so casually
..just reboot your datamover.


It actually says "Reboot or failover/failback the Data Mover." Some customers can reboot Data Movers during low I/O periods without needing change tickets. For those that can't, failover and failback essentially accomplishes the same thing.

6 Operator

 • 

8.6K Posts

November 11th, 2008 15:00

i love option two in that solution ..so casually ..just reboot your datamover.


I think a umount/mount with a fsck in between would serve the same purpose but a reboot is easier to explain

6 Operator

 • 

8.6K Posts

November 11th, 2008 15:00

I would still open a case with EMC support just to sure

I know that we have been moving more and more config file off .etc, but I'm not sure what for your specific config might have been there
For example the quotas definition file used to be in there

Easiest would be if you still have checkpoint of that file system so that you can see what was there

11 Legend

 • 

20.4K Posts

 • 

87.4K Points

November 11th, 2008 19:00

i hate rebooting, failing over and back anything when it comes to enterprise storage. I am not running a storage server here. I have high expectations of EMC as do my customers of me.

90 Posts

November 11th, 2008 21:00

For future use, I highly suggest not having any access to the root of volumes from external sources. I would create qtrees (can set quota or not) and point all shares and/or exports to a qtree to hide things like .etc

1 Rookie

 • 

13 Posts

November 13th, 2008 04:00

Amol,

Thanks for finding that support doc. I'm still really new to EMC products and finding things on Powerlink has been driving me crazy. It looks like this will help with the lost+found but doesn't talk about the .etc dir. I'm going to open a support ticket just in case. Thanks for the help.

Jarrod

1 Rookie

 • 

13 Posts

November 13th, 2008 04:00

Rainer,

I'm definitely going to call EMC on this one. It looks like emc66658 will resolve the problem with the lost+found dir but does not address .etc. Thanks for the help.

Jarrod

108 Posts

November 13th, 2008 06:00

Your Welcome !!!

Long time ago I was new too ;-)

Anyway, don't forget to post updates here once the issue is resolved. :-)

1 Rookie

 • 

13 Posts

November 13th, 2008 10:00

Well, I got off the phone with support. It turns out that while the lost+found dir can be recreated fairly easily, there is no way to recreate .etc. I was told to make a backup of the fs in question and delete it, then recreate it.

The lesson to learn here is protect the .etc folder in the root of your fs as much as you can!

Thanks to everyone for your help!

108 Posts

November 13th, 2008 12:00

That what I was thinking too. litle extra work for you........ ;-)

You can easily hide those files :-)
have a look --> emc140118
No Events found!

Top