NAKIVO Backup & Replication
Store NAKIVO Backup & Replication backups in XNS storage through a generic S3-compatible backup repository.
Overview
NAKIVO Backup & Replication backs up virtual machines (VMware, Hyper-V, Proxmox VE, Nutanix AHV), physical Windows and Linux machines, Microsoft 365 and file shares. It runs on Windows, Linux or a NAS. With XNS, NAKIVO writes its backups to a bucket on your Relayer: you add the Relayer to the NAKIVO Inventory as Generic S3-compatible Storage, then create a Backup Repository in a bucket there.
Not yet tested by XNS. Settings below follow the tool’s documentation and the gateway’s requirements.
XNS could not run NAKIVO itself: both the trial and the Free Edition download forms require a phone number. The bucket requirements below were checked against the Relayer directly: a scoped key with the canned ReadWrite policy can create a bucket with Object Lock and versioning on, but is refused the Object Lock calls NAKIVO lists for immutable backups.
Connect to XNS
Every S3 tool connects to your Relayer with the same settings. Enter them exactly as shown; the notes below the table cover the two that most tools get wrong by default.
| Setting | Value |
|---|---|
| Endpoint (HTTP) | http://<relayer-host>:9000, where <relayer-host> is the address of the machine running your Relayer. Always available. |
| Endpoint (HTTPS) | https://<relayer-host>:9443, available only after a certificate is installed on the Relayer. Port 9000 never serves HTTPS. |
| Region | us-east-1 |
| Addressing | Path-style. In tools that ask, turn on force path style or path-style access, and turn off virtual-hosted style. |
| Signature | Signature Version 4 (V4). Version 2 is refused. |
| Access key and secret key | For a tool, create a scoped key: IAM → Create IAM User in the Relayer UI. The secret is shown once, so copy it then. The install’s root pair is on Configure → Credentials; keep it for administration. |
| Bucket | Create it first: Files → Create Bucket in the Relayer UI, or let the tool create it. Names are 3-63 characters of lowercase letters, digits, hyphens and periods. |
HTTPS and certificates. The Relayer UI offers two certificates for port 9443: Let’s Encrypt, which needs a domain managed in Cloudflare and a Cloudflare API token, or self-signed. Most tools refuse a self-signed certificate unless certificate verification is turned off in the tool. On a private network, plain HTTP on port 9000 avoids the question.
Path-style. A tool left on virtual-hosted style (often labeled Amazon S3 or AWS) sends the bucket name in the hostname, and the Relayer cannot find the bucket. Choose the tool’s S3 compatible or generic S3 option and set path-style.
Not supported: server-side encryption with customer-provided keys (SSE-C) or KMS keys (SSE-KMS), bucket website hosting and bucket replication. Server-side encryption with AES256 is accepted. See Encryption at the Gateway, Identity and Access and Configure.
Before you start
- NAKIVO Backup & Replication 11, from nakivo.com. The 15-day trial and the Free Edition both need a name, email, phone number and company in the download form; no payment details.
- A scoped key for NAKIVO on your Relayer (see the table above). For immutable backups, the key also needs the Object Lock permissions listed under Known limits.
- A bucket used by NAKIVO only. NAKIVO’s documentation says the bucket must have Object Lock and versioning on, and that it must hold no other data. Object Lock can be turned on only when the bucket is created, so create the bucket with it, for example:
aws s3api create-bucket --bucket <bucket> --object-lock-enabled-for-bucket \ --endpoint-url http://<relayer-host>:9000 --region us-east-1
The Relayer turns versioning on with Object Lock. Bucket names must be 3-63 characters of lowercase letters, digits, hyphens and periods, start and end with a letter or digit, and not look like an IP address; NAKIVO does not list other buckets.
Step 1: Add the Relayer to the Inventory
- In the NAKIVO web interface, choose Settings → Inventory and click +.
- On the Platform page, choose Cloud Storage. On the Type page, choose Generic S3-compatible Storage, not Amazon S3.
- Fill in the Options page:
| Field | Value |
|---|---|
| Display name | A name for the Relayer, for example XNS storage. |
| Service endpoint | The full URL with the scheme: http://<relayer-host>:9000, or https://<relayer-host>:9443 when the Relayer has a certificate. |
| Region code | Leave it blank, as NAKIVO recommends, or enter us-east-1. The Relayer accepts either. |
| Access key ID or Username | The scoped key’s access key. |
| Secret access key or Password | The scoped key’s secret key. |
- Click Connect. For an HTTPS endpoint NAKIVO shows the Certificate Details; check them and click Accept. Click Finish.
NAKIVO uses path-style requests for generic S3-compatible storage (its API list for this feature includes path-style access), so there is no addressing setting to change.
Step 2: Create the backup repository
- Choose Settings → Repositories, click + and choose Create new backup repository.
- On the Type page, choose Cloud & S3-compatible Storage. On the Vendor page, choose Generic S3-Compatible Storage.
- On Name & Location, enter a Name, pick the Assigned transporter, select the XNS Account from Step 1 and the Bucket you created.
- On Options, choose the compression level under Storage Savings (it cannot be changed later), and set the verification and detach schedule if you want them. Click Finish.
Step 3: Back up and restore
Create a backup job as usual (Data Protection → + → Backup job) and pick the XNS repository as the destination. Forever-incremental backups are not available on S3-compatible repositories, so set periodic full backups in the job’s retention options. After the first run, test a restore: recover a few files or a whole machine to a new location and compare them with the originals.
Step 4: Remove old backups
NAKIVO deletes recovery points from the bucket according to the job’s retention settings. To remove a repository’s data, delete the backups in NAKIVO before removing the repository. Recovery points still under immutability cannot be deleted until their retention ends.
Known limits
- Not tested with NAKIVO. NAKIVO says only the S3-compatible vendors it has tested are supported, and XNS is not on that list. Its documentation adds that other vendors may work if they support the APIs it uses; the Relayer supports all of them (bucket listing and location, versioned listing, object read, write, copy and delete, multipart upload and upload-part-copy, and Object Lock retention).
- Immutable backups need more than the ReadWrite policy. Under the canned ReadWrite policy the Relayer refuses
GetObjectLockConfiguration,GetObjectRetentionandPutObjectRetentionwithAccessDenied. NAKIVO uses these calls for immutability, so give NAKIVO’s key the FullAccess policy, or a custom policy that addss3:GetObjectLockConfiguration,s3:GetObjectRetentionands3:PutObjectRetentionto ReadWrite, in IAM in the Relayer UI. Use these action names: the Relayer checkss3:GetObjectLockConfiguration, not the AWS names3:GetBucketObjectLockConfiguration. See Identity and Access. Whether NAKIVO accepts the repository without these calls is untested. - Object Lock must be set when the bucket is created. It cannot be turned on for an existing bucket. A scoped ReadWrite key can create the bucket with Object Lock on, but cannot change versioning afterward (
PutBucketVersioningis refused). - Port 9000 is plain HTTP. Use
http://with port 9000 andhttps://with port 9443. For a self-signed certificate on 9443, accept its details when NAKIVO shows them. - No forever-incremental backups on S3-compatible repositories; this is a NAKIVO limit for every S3 location.
- One bucket per repository, NAKIVO data only. Do not store other data in the bucket.
- Not supported by the Relayer: SSE-C and SSE-KMS server-side encryption, bucket website hosting and bucket replication. Use NAKIVO’s own backup encryption.
