Mounting Snapshot LUNS on the Same Host as the SRC LUNs (Linux)

This steps will surely help guys who are having dev/test environment and also will save lot of time then restoring the whole data from the backups

Preparing the DB for Snapshot

1. Shutdown or Begin Hot backup on the DB you want to snap

Create the Snapshot on the Storage

2. Create the Snapshot on the LUNs related to <VGNAME>

Change the PVID and VGID on the DB Host

3. Unmount the file systems in the VG

4. Deactivate the Original VG
a. vgchange -a n <VGNAME>

5. Change the UUID of the PVs(Disks) in the VG (Make sure to do it on all Disks in the VG)
b. pvchange -u <DISKSinVG>

6. Change the UUID of the Original VG
c. vgchange -u <VGNAME>

7. Rename the original VG
d. vgrename <VGNAME> <VGNAME>_orig

Map the Snapshot to the DB Host

8. Make sure the PVIDs and VGIDs of the Original VG are changed before presenting the Snapshot
LUN to the Host.

9. On the Storage System, export/map the new Snapshot LUN to the DB Host .

Importing the Snapshot VG to the Host

10. Scan for the new Snapshot LUN
e. echo "- - -" > /sys/class/scsi_host/<hostX>/scan

11. Create Multipath devices for Snapshot LUN
f. multipath

12. Scan for the PVs(It shouldn't show any duplicate UUIDs)
a. pvscan

13. Import the Snapshot VG
g. vgimport <VGNAME>

14. Rename the VG on the Snapshot LUN
a. vgrename <VGNAME> <VGNAME>_Snap1

15. Generate a new PVID for Snapshot LUNs
a. pvchange -u <DISKSinSNAPVG>

16. Generate a new VGID for the Snapshot VG
a. vgchange -u <VGNAME>_Snap1

17. Generate a new VGID for Snapshot VG

18. Rename the Original VG to its actual name
b. vgrename <VGNAME>_orig <VGNAME>

19. Activate Original VG as well as Snapshot VG
a. vgchange -a y <VGNAME>
b. vgchange -a y <VGNAME>_Snap1

20. Add FSTAB Entries for the Snapshot VG

21. Mount the file systems in Original VG as well as the Snapshot VG.

Unified share in VNX

Today I would like to share something which i was not able to find on the manuals, I got a request to share a filesystem on our VNX so that the users using Windows and Linux clients can access the share.

I was searching for ideas and with the help of one of my colleague was able to find a way to do it. So, in order to do this we have to create a Virtual Datamover which will server the data for our Windows client and for the NFS request from Linux clients.

I really liked the concept of EMC Celerra to create a Virtual Datamover (VDMs) which adds more security by letting the clients access the filesystem from VDMs ip and not populate of the physical DMs.

Lets just create a new VDM

[nasadmin@cs0 ~]$nas_server -name VDM01 -type vdm -create server_2

[nasadmin@cs0 ~]$ nas_server -l -a
id      type  acl  slot groupID  state  name
1        1    0     2              0    server_2
2        4    0     3              0    server_3

id       acl  server    mountedfs       rootfs  name
8        0    1         1               1       VDM01

We will create a new pool for this purpose and a new shared filesystem 

[nasadmin@cs0 ~]$nas_pool -create -name clar_r5_media -volumes d7,d8

[nasadmin@cs0 ~]$nas_fs -name unify_fs -create size=800G pool=clar_r5_media

Here we are creating a CIFS server under the VDM and adding it to domain which will let user which are part of the active directory access the filesystem

[nasadmin@cs0 ~]$server_cifs VDM01 -Join compname=cifs01,domain=ad.test.com,admin=Administrator,ou="ou=Computers:ou=EMC Celerra"

There are three simple steps which we have to do once the filesystem is created, creating a mountpoint, mounting the filesystem and exporting it to the clients

Lets just do it for windows clients which will be using CIFS protocol access the files

[nasadmin@cs0 ~]$server_mountpoint VDM01 -create /unify_fs

[nasadmin@cs0 ~]$server_mount VDM01 -option accesspolicy=MIXED_COMPAT unify_fs /unify_fs

[nasadmin@cs0 ~]$server_export VDM01 -Protocol cifs -name unify_fs /unify_fs

Now, lets just do it for Linux clients which will be accessing the share using NFS but before we do that we have to find the root path of the VDM so that we can add it with the shared filesystem

[nasadmin@hct-cs0 ~]$ nas_fs -l |grep VDM01
24        y    1   0     157       root_fs_vdm_VDM01   1

[nasadmin@cs0 ~]$ server_df server_2 root_fs_vdm_VDM01
server_2 :
Filesystem          kbytes         used        avail capacity Mounted on
root_fs_vdm_VDM01
                1907824776         8576   1907816200    0%    /root_vdm_8/.etc

Here's another way to find that

[nasadmin@cs0 ~]$ server_mount server_2 |grep VDM01
root_fs_vdm_VDM01 on /root_vdm_8/.etc uxfs,perm,rw

Now, we can create mountpoint, mount the share and export

[nasadmin@cs0 ~]$server_mountpoint server_2 -create  /root_vdm_8/unify_fs

[nasadmin@cs0 ~]$server_mount VDM01 -option accesspolicy=MIXED_COMPAT unify_fs /root_vdm_8/unify_fs

[nasadmin@cs0 ~]$server_export server_2 -Protocol nfs -name unify_fs /root_vdm_8/unify_fs


[nasadmin@cs0 ~]$ server_export VDM01
VDM01 :
share "unify_fs" "/unify_fs" type=Global umask=022 maxusr=4294967295 netbios=CIFS01


[nasadmin@cs0 ~]$ server_export server_2
server_2 :
export "/root_vdm_8/unify_fs" name="/unify_fs"


That was pretty easy!!