I need to run snapmirror for a 8TB volume from site A to site B through WAN. The initial effort to get the snapmirror initialization via WAN failed, it tooks more than 3 weeks to run baseline and yet still failing .  Here comes SnapMirror to tape (smtape)  to save me from wasting my time praying so the initialization will get through.  I have my Oracle StorageTek SL150 installed at site B recently and here comes the challenges.

My company is using EMC Networker 8 , i have experience with Legato Networker(before it was acquired by EMC) , not so friendly. I was then introduced to Commvault ,  it could be very user friendly but due to it flexibility, it is sometimes complicated too.   There are 2 ways for me to accomplish my mission – export the data to tape , send it to another site, restore tape data, snapmirror resync !

1. Simple way – could apply to any backup software

2. Networker way

Part 1 is covering “simple way”

Ugly screenshot

simple way

There are 6 steps involved:

1.smtape to a source temp volume

2. Backup source temp volume to up to your tape using backup software

3. Send the tape to the destination site

4. Restore the date from tape to destination temp volume

5. smtape restore from temp volume to exact volume

6. snapmirror resync to establish the connection.

 
Example output :
1. create temp volume and smtape to it

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
filer_A> vol create smtape_test_source -l en_US aggr0 50g
Creation of volume 'smtape_test_source' with size 50g on containing aggregate
'aggr0' has completed.
 
filer_A> qtree create /vol/smtape_test_source/source
filer_A> priv set diag
Warning: These diagnostic commands are for use by NetApp
         personnel only.
filer_A*> smtape backup /vol/VMware_ISO /vol/smtape_test_source/source/dump_file
Job 13 started.
filer_A*> smtape status
Job ID Seq No Type    Status      Path                   Device                 Progress
    13      0 Backup  Active      /vol/VMware_ISO        /vol/smtape_test_source/source/dump_file 230.676 MB
filer_A*> smtape status
job not found

2. [Backup using your backup software to tape]
3. [send the tape to destination]

4. Retore the tape data to temp volume. I am using Networker “recover” command to restore the data from tape to filer.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
filer_B> vol create smtape_test_dest -l en_US aggr0 60g
 
[root@backup_server]# recover -R filer_B -c filer_A -d /vol/smtape_test_dest /vol/smtape_test_source
Current working directory is /vol/smtape_test_source/
recover> dir
total 256
03/11/13 17:24            <DIR>         source
recover> pwd
/vol/smtape_test_source/
recover> list
0 file(s) marked for recovery
recover> destination
recover files into /vol/smtape_test_dest
recover> ls
source
recover> cd source
recover> ls
dump_file
recover> cd ..
recover> ls -l
total 256
drwxrwxrwx root             4096 Mar 11 17:24 source
recover> add source
2 file(s) marked for recovery
recover> recover
Recovering 2 files within /vol/smtape_test_source/ into /vol/smtape_test_dest
Volumes needed (all on-line):
        A00035 at /dev/rmt/0cbn
Total estimated disk space needed for recover is 2113 MB.
Requesting 2 file(s), this may take a while...
42795:nsrndmp_recover: Performing recover from Non-NDMP type of device
Host = backup_server.example.com (192.168.1.123) port = 9661
42689:nsrndmp_recover: Performing DAR Recovery..
42617:nsrndmp_recover: NDMP Service Log: DIRECT ACCESS RECOVERY (DAR) requested
 
42937:nsrdsa_recover: Performing Immediate recover
42940:nsrdsa_recover: Reading Data...
42617:nsrndmp_recover: NDMP Service Log: RESTORE: RESTORE IS DONE
 
42942:nsrdsa_recover: Reading data...DONE.
42927:nsrndmp_recover: Successfully done

5. smtape command to restore the data from temp volume to the snapmirror destination volume.

1
2
3
4
5
6
7
8
9
10
11
12
filer_B*> vol create test_restore -l en_US aggr0 60g
Creation of volume 'test_restore' with size 60g on containing aggregate
'aggr0' has completed.
 
filer_B*> vol restrict /vol/test_restore
Volume 'test_restore' is now restricted.
 
filer_B*> smtape restore /vol/test_restore /vol/smtape_test_dest/source/dump_file
 
Job 9 started.filer_B*> smtape status
Job ID Seq No Type    Status      Path                   Device                 Progress
    11      0 Restore Active      /vol/test_restore      /vol/smtape_test_dest/source/dump_file 430.848 MB

6. Once the restoration completed, check on snapmirror status to see if the connection ready. Resync one you see it!

1
2
3
4
filer_B*> snapmirror status
Snapmirror is on.
Source                                                      Destination                                State          Lag        Status
snapshot_for_smtape.8add9a0e-8a2d-11e2-a408-123478563412.0  filer_B:test_restore                      Snapmirrored   21:08:21   Idle
1
2
3
4
5
6
7
8
filer_B*> snapmirror resync -S filer_A:VMware_ISO filer_B:test_restore
NOTE: Destination volume test_restore is already a replica.
NOTE: Resync will not need to revert the volume.
The resync base snapshot will be: snapshot_for_smtape.8add9a0e-8a2d-11e2-a408-123478563412.0
Are you sure you want to resync the volume? y
Tue Mar 12 14:35:17 MYT [filer_B:replication.dst.resync.success:notice]: SnapMirror resync of test_restore to filer_A:VMware_ISO was successful.
Transfer started.
Monitor progress with 'snapmirror status' or the snapmirror log.

The drawback for this method is: require 2x of space in destination during the restoration. This is simple but space consuming. It is good if you have the extra space in your destination filer.

Part II will cover the restoration using Networker 8.6P4.