Backup is one of the primary usecase of Nutanix Objects. And we have certified all the top backup softwares with Objects which supports S3 as target storage. In this blog we will go over the migrating your backup workload to Nutanix Objects cluster.
For my tests, I have taken one of our partners, Commvault as the backup software.
Customer who may have been using Commvault to backup data to a cloud service such as Amazon AWS S3. The steps below describe how such a customer can bring that data back to on-prem.
While I have taken Commvault as an example, this would work with any application that could use object store with S3 APIs.
We are going to :
- Take a quick look at setup.
- Configure replication between Amazon S3 -> Nutanix Objects
- Change Commvault configuration to point to Objects endpoint.
- Generate data loss and recover data from backup replicated using Objects Replicator.
My test Commvault configuration :
In my test environment, I have Commvault configured to backup one VM to Amazon Aws S3.
- Commvault storage library configured with Aws S3 bucket : “backup-ups-vms”
- Commvault subclient protecting scalcia-project-vpx VM which is being backed up to Aws s3://backup-ups-vms.
On successful configuration, initiate one full backup to Aws s3://backup-ups-vms bucket.
In the above screenshot you can see, s3://backup-ups-vms has quite a lot objects. These all objects are pushed by Commvault . During backup commvault divides entire VM in smaller chunks 8MB/32MB (size may vary) . This VM is around 8.5GB in size and total objects created by Commvault is around 295. For more details on Commvault backup refer this post.
Configure Objects replicator to replicate Commvault backup to Nutanix Objects cluster :
I have appended additional config in the same replication1.json config file which we created in previous post.
This is how our latest config looks like now :
| { "replicator": { "LocalConfig1": { "Source": { "endpoint": "calculus.scalcia.com", "access_key_id": "ILOX_dH87avkdbjOagIvdiJ19j0HpLsy", "secret_access_key": "GoCdPuca8VP066btzwG6cnswa-YPUail" }, "Destination": { "endpoint": "numbers.scalcia.com", "access_key_id": "PJdSxU1AcbApjF_S79k5XDm3CbwmD4OV", "secret_access_key": "ZyD7hP01JkzOE-QntoaEBi777-useHyK" }, "Bucket_Map": { "vdi-bucket": [ "vdi-bucket-dst" ] } }, "AwsConfig1": { "Source": { "endpoint": "https://s3.amazonaws.com", "access_key_id": "XXXxxxXXXxxxx", "secret_access_key": "XXXxxxZZZZzzzzYYYYyyyy" }, "Destination": { "endpoint": "numbers.scalcia.com", "access_key_id": "PJdSxU1AcbApjF_S79k5XDm3CbwmD4OV", "secret_access_key": "ZyD7hP01JkzOE-QntoaEBi777-useHyK" }, "Bucket_Map": { "backup-vpx-vms": [ "local-backup-repo" ] } } }, "frequency": "1 h" } |
- LocalConfig1 : Replicates data between two Objects Instance.
- AwsConfig1 : Objects Replicator will replicate Aws s3://backup-ups-vms data -> Objects https://numbers.scalcia.com/local-backup-repo bucket
Quick look at both buckets before Starting Replication :
On left I am connected to Aws s3://backup-vpx-vms which has data backed up by Commvault. And on the right side we are connected to https://numbers.scalcia.com to s3://local-backup-repo bucket , which is empty .
Start Objects replicator service to initiate full replication :
| nutanix@NTNX-ORVM: ~ $objects_replicator_start -p replication1.json Checking if /home/nutanix/replication1.json exists. Using config present at /home/nutanix/replication1.json to start the replication. Setting STATS_TIMER to 2 seconds. Created symlink from /etc/systemd/system/multi-user.target.wants/supervisord.service to /usr/lib/systemd/system/supervisord.service. nutanix@NTNX-ORVM: ~ $ |
- After replication is started, wait until all the objects are replicated. Check Objects Replicators log for successful replication confirmation.
- And this is how both the buckets looks like now:
Both the buckets are synced and you should see bucket structure and all objects exactly same in both buckets.
(Note : This is only for test purpose, refer commvault official documentation for making any changes to Commvault settings.)
Proceed with next Test:
Change the Commvault configuration to point storage library to Nutanix Objects S3 bucket.
Login to Commvault Commcell
- Goto Storage Libraries
- Open Aws storage library settings
- Change Service Host to “http://numbers.scalcia.com” (i.e your on-prem Nutanix Objects S3 endpoint)
- Change Bucket to “local-backup-repo”
- Add local IAM credentials.
- Click on Verify button. This should validate settings successfully.
- Save settings. This will ask for confirmation, hit Yes.
So far so good, check Commvault logs for any error or warning.