VMware Solutions Discussions

How to SnapVault SMVI (VMware based) backups with OnCommand Core 5.0 and Host Agent 1.1 without SV-SMVI Script

charlesadams
12,003 Views

First and foremost, this is a rough draft and has the potential for revision. Screenshots will be added later. Please comment on this for improvements

Assumptions:

1) OnCommand Core 5.0 is installed using the Custom install method (not building canned options)

2) OnCommand Host Agent 1.1 is installed and registered with vCenter and OnCommand Core. (Can see Virtual machines as well as hosts, etc)

3) Needs a resource pool established for destination location.

4) All NetApp Controllers are licensed for SnapVault and SMVI, as well as had the SnapVault enabled via CLI.

5) All NetApp Controllers involved are registered with OnCommand Core.

6) Have an appropriate Protection policy setup with schedules, and the type of protection would be based off the "Back up" policy type. (Local backup on primary, then back up primary to secondary; I created one based off "Back up" and modified the schedules, retention, and called it "SMVI Backups")

Process:

Create Storage Service:

1) Open NetApp Management Console 3.1 and go to "Manage Data", then "Policies", then "Storage Services"

2) Click "Add", and select "Next".

3) Type in appropriate information, with a minimum of "Name" (I called it "Thin Provision NFS then SnapVault") and select "Next".

4) Select appropriate "Group" (I used Global), then Select "Next"

5) Select your appropriate "Protection Policy" (i.e. "SMVI Backups") and select Next.

6) For "Primary Data" I left defaults (Provisioning= None, vFiler = None, No resource pools selected). Select "Next"

7) Select destination "Resource Pool" and include it in the "Resource pools in this node" , leave the rest blank. Select "Next"

😎 Select "Finish", Close Management Console.

Create Policy:

1) Open OnCommand Core web console

2) Select "Policies" Tab.

3) Select "Create" -> "VMware Policy"

4) Create a Unique name (I used "VMware Remote Office")

5) Create a Schedule (I just created a daily schedule that runs at 10PM everyday.

6) Under "Backup Options" column I clicked "View" and checked the box for "Start remote backup after local backup"

7) Clicked "Ok"

Assign Policy:

1) Click "Server" tab.

2) Select "VMware VMs"

3) Select Virtual Machine to protect, right click and select "Back Up" -> then "Using New Dataset"

4) Assign a Name (I called mine SMVI Replicated) and Details, select "Data" and include any Virtual Machines to be replicated.

5) Under "Local Policy" I selected "VMware Remote Office". If you expand the policy you will see your schedules and options configured previously.

6) Under the Storage Service area click the drop down and select the Storage Service you created in the Management Console (i.e. Thin Provision NFS then SnapVault)

7) Click "Ok" at the bottom.

😎 Right Click the machine and select "Back Up Now"

9) Dataset name should be the once you just created (i.e. SMVI Replicated)

10) Retain for however long you want to keep it for. Check the box at the bottom that says "Start remote backup after local backup". Click ok.

This should start the process for conformance and creating a Volume and Qtrees on the destination. Once you run the backup you see via the CLI on the destination controller the SnapVault relationship via "snapvault status".

10 REPLIES 10

begasoftfiler
11,985 Views

Thanks. Helps me a lot.

Question:

When i create the Protection Policy (Step 6 - "Assumptions") i define schedules  for "Primary Data" Backup and "Primary data to Backup". Then again (Step 5 - "Create a Schedule") i create another scheduling at the Oncommand Level? What does that mean? Why there is another scheduling at the "oncommand" Level and what happens then...i mean we have already the scheduling at the Protection Policy which makes the Backup and then the Snapvault...?

Best Regards

begasoftfiler
11,985 Views

Thanks. Helps me a lot.

Question:

When i create the Protection Policy (Step 6 - "Assumptions") i define schedules  for "Primary Data" Backup and "Primary data to Backup". Then again (Step 5 - "Create a Schedule") i create another scheduling at the Oncommand Level? What does that mean? Why there is another scheduling at the "oncommand" Level and what happens then...i mean we have already the scheduling at the Protection Policy which makes the Backup and then the Snapvault...?

Best Regards

begasoftfiler
11,985 Views

Thanks. Helps me a lot.

Question:

When i create the Protection Policy (Step 6 - "Assumptions") i define schedules  for "Primary Data" Backup and "Primary data to Backup". Then again (Step 5 - "Create a Schedule") i create another scheduling at the Oncommand Level? What does that mean? Why there is another scheduling at the "oncommand" Level and what happens then...i mean we have already the scheduling at the Protection Policy which makes the Backup and then the Snapvault...?

Best Regards

nlfmmeijburgienasb
11,985 Views

Great article,

All works great, but the only relation it makes is a qtree snapmirror relationship. Is there a setting to use snapvault as a relation?

Also second question, you create a schedule for snapping a vm or datastore, and check the box "start remote backup after local backup" But in the protection policy (assumptions nr6) you also create a schedule for normal primary to backup with retentions, but for us it is important to keep the vmware consistent snapshots on the secondary volume, how do i set the retention for this consistent snapshots?

adaikkap
11,985 Views

What is the protection policy that your storage service is using ?

Is SV primary and SV secondary properly licensed on the source and destination Controller ?

Is dedupe enabled on the secondary volume if so what type of dedupe ? Is it scheduled, automated or On-Demand ? Only On-Demand dedupe will create SV relationships others will create QSM.

Regards

adai

nlfmmeijburgienasb
11,985 Views

Thanks for your quick reply, protection policy says nothing about snapvault, licence is in place, i have many snapvault relationships allready. dedupe is not enabled on the secondory volumes because i use thin provisioned volume with sizes of 4TB. And for the storage controller (IBM N3600, ontap 7.3.6) 2TB is maximum for dedupe. So you say i must enable de-dupe?

adaikkap
11,986 Views

So, dedupe is not the issue, as its not enabled. What is the protection policy that you are using ? in your storage service ?

Regards

adai

nlfmmeijburgienasb
11,986 Views

The protection policy connected to the storage service is a copy of the default "backup" policy, with daily at 23:00 and hourly as primary. and as backup schedule First Sunday at 8:00 plus weekly and daily. Or should i cancel al this backup schedules and only use the ones in create with the "virtual objects?"

I do not have one dataset which uses snapvault, only the ones i manualy created and imported (not virtual).

The storage service is connected to a NFS volume

ravindhar
11,986 Views

One last thing can you please check dfm option list pmQSMBackupPreferred is set to "No"

laniocedric
7,179 Views

This article helps me a lot to backup my vmware vm.

But I have a question. I manage a dataset for one vm  to test, and I find a sort of issue : my secondary volume, wich is automatically create when I confirm my dataset is more than the double of size of my primary volume.

In my example, I create a local backup of a vm with wmware consistent backup and remote backup after. This vm is in a 100 GB volume on the primary and the volume create by the dataset is a thin provisionned volume of 249 GB.

I find this in the logs of jobs of the dataset :

Dynamic secondary volume sizing is enabled.

by the way, it does that after :

DSS: Secondary volume FASXXX2:/SV_vol_srvweb_vm (885): current total=255868MB,

current used=76994MB, new total=232705MB, volume margin=10, adjusted new

total=255975MB, resize_up_only=0. Secondary Volume Size Limits: Opt DSS Max=0MB,

Limit to Aggr ("No")=13733114MB, Dedupe Enabled ("No")=3145728MB, Max Vol

Limit=15069184MB.

When I search in the storage service i create, when i look at the secondary provisionning policy, it is default. (I assume thats for point i need to modify ?) but I can't find or define a secondary policy that seems to be convenient (only setup dedup).

If someone has an answer to this, you're very welcome

Public