Machines replication

Machine replication consists of user-initiated and system-managed steps.

User Actions

1. Create a machine group to organize replicated machines.

2. Install and configure replication agents, including agent type selection and connection settings.

3. Configure replication settings, such as replication schedules and retention policies.

4. Start Protection to create the initial consistent machine snapshot.

System Actions

5. Retrieve machine snapshots and detect changes between consecutive snapshots.

6. Transfer changed data to the target storage using WAN optimization and deduplication technologies.

7. Store incremental backups according to the configured snapshot retention policies.

Create a machine group

Machine groups help organize replicated machines and simplify protection management. Each group can contain machines that belong to the same environment, department, customer, application, or disaster recovery scenario.

Once a customer dashboard has been created as part of product's initial configuration for replicating machines, navigate to the main page and create machine group.

Use Cases

• Environment-based organization

Create separate groups for development, test, and production environments to manage replication settings independently.

• Application grouping

Place all machines that belong to the same application or service in a single group to simplify protection and recovery operations.

• Department-based management

Organize machines by business unit, such as Finance, HR, Engineering, or Sales, and delegate management responsibilities accordingly.

• Recovery planning

Group machines that participate in the same Recovery Plan to simplify failover and recovery operations.

• Geographical organization

Create groups for different sites, offices, or regions to manage replication according to location-specific requirements.

Getting started with replication agents

Note

The replication agent requires a virtual machine with at least 2 vCPUs and 4 GB of RAM.

1. Download the replication agent in one of the following ways:

• On the customer page, select Protect New Platform from the machine group's Actions menu. For the complete workflow, see Protect New Platform.

• On the customer page, select Manage Clouds, locate the source cloud, and select Download Replication Agent from the cloud's Actions menu. For more information, see Manage Source and Failback Clouds.

• Select Download Agents in the main menu.

2. Follow the wizard instructions:

• Step 1. (Optional) Select a customer.

• Step 2. Select the agent type:

  • External agents: VMware, OpenStack, and oVirt.
  • Internal agents: Windows and Linux (installed directly on the protected machine), and the PostgreSQL backup agent (available only for the Storage cloud type).

• Step 3. Configure the agent settings. The available settings depend on the selected agent type and are described in the corresponding agent installation section. See Replication agent settings for details. To have detected machines automatically assigned to a group, select a previously created machine group.

• Step 4. Download and install the agent.

Use case: add machines to a specific group

Note

To add machines to a specific group, a Control Panel administrator must first create the group.

1. After the group is created, open the group's ... menu and select Protect New Platform

2. Select a supported agent type.

3. Configure the agent settings: select a previously created group and specify the required parameters. The available parameters depend on the selected agent type. For details, see the corresponding agent configuration section:

• VMware Agent
• oVirt Agent
• OpenStack Agent
• Linux Agent
• Windows Agent
• PostgreSQL backup agent

4. In Step 4, select Download agent and follow the generated deployment instructions.

After the agent is downloaded and installed, all detected machines automatically appear in the machine group selected in the previous step. The machines are initially displayed in the Unprotected state.

Replication agent settings

VMware Agent

Client for VMware (VMware agent) - a Linux machine, supplied as an OVA file, that is deployed to each ESXi host or cluster and that replicates machines on these hosts. Standard VMware tools are used to detect changes and retrieve deltas on virtual machines via the Changed Block Tracking (CBT) API. These tools, in turn, trigger the operating system's mechanisms (such as quiesce) to ensure data consistency within the replicas.

An agent instance must have at least 2 vCPUs and 4 GB of RAM and must be deployed on each ESXi host or cluster containing machines designated for replication.

Note

An ESXi host or cluster detection depends on cloud configuration settings according to the discovery level.

Requirements

To ensure the replication process is successful, it is necessary to set the appropriate permissions. The permissions can be set at a high level of the vCenter hierarchy or at least at the host/cluster level where the replication agent is deployed.

Note

If you set permissions at the host/cluster level, you should also add a read-only permission (empty role) at the datacenter level within the vCenter hierarchy. This permission should be set to propagate to children for HotAdd to work or without propagation for NBD mode, though some functionality may be limited (e.g., retrieving networks, storage information such as free space, etc.).

Additionally, set permissions if you want to use HotAdd transport mode.

VMware Replication Agent requires the following user permissions in vSphere (Role "VMware Consolidated Backup user" in vCenter):

• Datastore - Browse
• Datastore - FileManagement
• Virtual machine - Change Configuration -- Disc Lease
• Virtual machine - Interact -- Power On
• Virtual machine - Interact -- Power Off
• Virtual machine - Provisioning -- Allow read-only disk access
• Virtual machine - Provisioning -- Allow virtual machine files upload
• Virtual machine - Provisioning -- Allow virtual machine download
• Virtual machine - Snapshot management -- Create snapshot
• Virtual machine - Snapshot management -- Remove Snapshot

Permission to access CBT is necessary for the correct performance of the application. To enable CBT:

Virtual machine - Change Configuration - Disc change tracking

Additionally, it is recommended to include the following global permissions:

• Global - Disable methods
• Global - Enable methods
• Global - Licenses

A list of permissions for vCenter version 8.x:

Table 1: A list of permissions for vCenter version 8.x.

Privilege/Permission Name in the API	Privilege/Permission Name in the vSphere Client	Description
Datastore.Browse	Browse datastore	Allows browsing files on a datastore.
Datastore.FileManagement	Low level file operations	Allows performing read, write, delete, and rename operations in the datastore browser.
VirtualMachine.Config.DiskLease	Change Configuration → Acquire disk lease	Allows disk lease operations for a virtual machine.
VirtualMachine.Interact.PowerOn	Interaction → Power On	Allows powering on a powered-off virtual machine, and resuming a suspended virtual machine.
VirtualMachine.Interact.PowerOff	Interaction → Power Off	Allows powering off a powered-on virtual machine. This operation powers down the guest operating system.
VirtualMachine.Provisioning.DiskRandomRead	Provisioning→Allow read-only disk access	Allows opening a disk on a virtual machine for random read access. Used mostly for remote disk mounting.
VirtualMachine.Provisioning.PutVmFiles	Provisioning →Allow virtual machine files upload	Allows write operations on files associated with a virtual machine, including vmx, disks, logs, and nvram.
VirtualMachine.Provisioning.GetVmFiles	Provisioning →Allow virtual machine download	Allows read operations on files associated with a virtual machine, including vmx, disks, logs, and nvram.
VirtualMachine.State.CreateSnapshot	Snapshot management → Create snapshot	Allows creation of a snapshot from the virtual machine’s current state.
VirtualMachine.State.RemoveSnapshot	Snapshot management → Remove Snapshot	Allows removal of a snapshot from the snapshot history.
VirtualMachine.Config.ChangeTracking	Change Configuration → Toggle disk change tracking	Allows activating or deactivating of change tracking for the virtual machine's disks. Switching it on is necessary to access CBT.
Global.DisableMethods	Disable methods	Allows servers for vCenter Server extensions to deactivate certain operations on objects managed by vCenter Server. We highly recommend turning it on.
Global.EnableMethods	Enable methods	Allows servers for vCenter Server extensions to activate certain operations on objects managed by vCenter Server. We highly recommend turning it on.
Global.Licenses	Licenses	Allows viewing installed licenses and adding or removing licenses. We highly recommend turning it on.

Note

In case of using vCloud, vCenter user requires one extra permission to operate: Profile-driven storage → Profile-driven storage view

• Ports for correct agent work:

  • DR host - tcp/80, tcp/443

  • vSphere host - tcp/443

  • ESXi host(s) - tcp/udp/902

  • Send logs to the Acura cluster - udp/12201

VMware Replication Agent uses VMware snapshots and VMware CBT API in order to create consistent replicas of machines' data.

This implies the following considerations regarding the VMware storage:

• VMware snapshots consume storage to retain copy-on-write buffer, so it is recommended to have at least 10% free space available on VMware storage.
• VMware puts additional load on storage while creating snapshots or running machines with existing snapshots.

Please consider that storage performance warning thresholds need to be adjusted in order to meet this increased load during replication.

Warning

The source machine must have VMware Tools installed manually prior to any replication procedures for it to display its network information in the target VMware ESXi correctly.

Permissions for HotAdd

Add permissions for using HotAdd transport mode either at the hierarchy level where the agent is deployed or directly on the agent itself, ensuring they are at the required level or higher:

• Virtual machine - Change Configuration -- Add existing disk

• Virtual machine - Change Configuration -- Advanced configuration

• Virtual machine - Change Configuration -- Remove disk

Installation

The first two steps that must be completed to successfully install the VMware replication agent are described in the section Machines replication and distribution between groups.

Step 3. Agent options

Select the machine group that will contain the protected machines, configure the VMware cloud access settings, and select Next. To create a group first, see Create a machine group.

If the platform is not listed, select New VMware vSphere instead of Registered VMware vSphere, complete all fields, and select Next. Use the ? icons for field-specific help. The wizard validates the values before opening Step 4.

Table 2: New VMware vSphere fields description

Field	Description	Example
Cloud name	The name of the cloud which will be shown in the UI. The name must be unique.	cloud
Endpoint	Endpoint to connect to	cloud.example.org
Login	User login	username
Password	Password to access the target cloud	password
Use enforce quiesce strategy	By default, the agent will try to quiesce the machine, create non-quiesce snapshot in case if quiesce is falling. When the checkbox is checked, non-quiesce snapshot will not be created.	unchecked
Use cluster level discover	Enable devices discover on cluster level. If unchecked, devices discover will be performed on host level. Details.	unchecked
On replication start	Whether to change source machine power state on replication start. Details.	No action
On replication finish	Whether to change source machine power state on replication finish. Details.	No action
Storage threshold per machine disk	Run application on if datastore has free space X percent of total disk size (0 – unlimited)	0
Free space storage threshold	Run application on if datastore has X percent of total disk size (0 – unlimited)	0
VDDK Compression	Use VDDK compression. Select the option from the list: No compression, zlib, fastlz, skipz.	No compression
Host/Cluster name	Name of the Host or Cluster (multiple allowed, separated by commas). Enables advanced discover mode. This field is optional. Details.	Empty
VDDK transport mode	Select the VDDK transport mode.	NBD:NBDSSL
Parallel disks uploading	Enable parallel disk uploading. By default, source disks will be uploaded sequentially. If checked, source disks will be uploaded in parallel. Enabling this setting increases the RAM size requirements for the replication agent.	unchecked
Filter VMs by name prefixes	Check to discover only VMs with names that match the provided prefixes. Details.	unchecked
Name prefixes filter	Active only when the ‘Filter VMs by name prefixes’ option is enabled. A comma-separated list of name prefixes. Details.	WIN_,TEST_
Filter VMs by custom attributes	Check to discover only VMs that match the provided custom attributes. Note that custom attributes are not available on ESXi: vCenter server endpoint is required. Details.	unchecked
Custom attributes filter	Active only when the ‘Filter VMs by custom attributes’ option is enabled. A comma-separated list of custom attributes in key=value format. Details.	TYPE=PROD,TYPE=TEST

Machine discovery levels

In the source cloud or replication agent settings, you can configure the level at which virtual machines are discovered for protection. To do this, two fields are provided: Use cluster level discover and Host/Cluster name.

The machine discovery level is determined by the Use cluster level discover setting. By default, this setting is disabled, which means that machine discovery is performed at the host level. When enabled, discovery is performed at the cluster level.

In the Host/Cluster name field, specify the hosts or clusters within which the search is performed. Rules:

• Leave the field empty if only machines from the host or cluster where the replication agent is deployed should be included in the protection list.

• If necessary, specify multiple hosts or clusters; use a comma as a separator.

• Do not assign the same hosts or clusters to different replication agents.

• To perform discovery and replication, the user specified in the source cloud settings must have sufficient permissions. Learn more.

Fields On replication start and On replication finish

The On replication start and On replication finish settings allow you to power off or power on the source virtual machine when replication starts/ends. The default value is No action: the state of the machine will not be changed.

Filter fields

Use Filter VMs by name prefixes and Filter VMs by custom attributes to filter machines. These filters are useful when you need to limit the number of replicating devices. You can use them either together or separately.

Typical cases and rules for using filters:

• If you need to discover VMs with multiple prefixes, you can specify them as a comma-separated string.

  

• If no filters are used (both checkboxes are unchecked), all VMs will be discovered according to the other settings.

• If at least one filter is enabled but its input field is empty, no VMs will be discovered.

• Filters cannot contain special characters such as commas (","), slashes ("/"), etc., and must not end with spaces.

• Make sure to set keys and values correctly to get the expected result.

• In the case of custom attributes, an exact match of both keys and values is required.

• If both filters (prefixes and custom attributes) are set, only VMs that meet both criteria (i.e., in the intersection of the filtered sets) will be discovered.

• The replication agent VMware retrieves updated settings for the corresponding source cloud on the next heartbeat after saving (approximately every 30 seconds by default).

To enable VM filtering by attributes, set them for the source VMs in vCenter:

1. Open the VM page.

2. Go to Actions → Tags & Custom Attributes → Edit Custom Attributes.

3. If the attribute already exists, set or change its value. Otherwise, add a new attribute (type: Virtual machine) by specifying its name and value.

Static settings

Additionally, it is possible to specify static network settings. Set Agent network settings to Static and fill in the fields.

Table 3: Agent network settings static fields description.

Field	Description	Example
Agent IP	The IP address of the agent	172.22.6.230
Subnet mask	The subnet mask of the agent	255.255.0.0
Gateway	Agent gateway	172.22.1.2
DNS servers	DNS of the agent’s server. You can specify multiply addresses separated by commas.	1.1.1.1, 2.2.2.2

Step 4. Protection instructions

Step 4 displays Download agent and the deployment instructions:

1. Select Download agent to download the OVA file, and deploy it on each ESXi host in your VMware cluster you want to protect.
2. Launch deployed VMs (agents) to protect machines on each host.

Machines will be detected and shown in a target group in a few minutes after agent deployment and launch. The detected machines appear in the selected group with the Unprotected status. Select the machines and choose Actions → Start Protection.

Note

VMware vSphere can issue a warning about the presence of an unknown configuration parameter for the virtual machine. The service parameter "hvragent" is added deliberately and it is not a security risk.

VMware Agent configuration

Use the shared external-agent TUI configuration. VMware agent logs are stored in /var/log/cabrio/cabrio.log.

oVirt Agent

oVirt Agent is an external replication agent type that is deployed as a separate instance in the source environment.

Requirements

Note

Replication Agent for oVirt requires superuser permissions for installation.

• Hardware:

  • Memory: at least 4 GB RAM

  • CPU: at least 2 vCPUs

  • Disk space: at least 10 GB required. Disk will be downloaded with fix size = 10Gb. It is whole disk of new deployed VM.

• Ports for correct agent work:

  • DR host - tcp/443

  • Send logs to the Acura cluster - udp/12201

An agent instance must be deployed on each project that has machines intended for replication.

Installation

The first two steps that must be completed to successfully install the oVirt replication agent are described in the section Machines replication and distribution between groups.

Step 3. Agent settings

Select the machine group that will contain the protected machines, configure the oVirt cloud access settings, and select Next. To create a group first, see Create a machine group.

If the platform is not listed, select New oVirt instead of Registered oVirt, complete all fields, and select Next. Use the ? icons for field-specific help. The wizard validates the values before opening Step 4.

Table 4: New oVirt fields description.

Field	Description	Example
Cloud name	The name of the cloud which will be shown in UI. The name must be unique	ovirt_source
oVirt API endpoint	oVirt API endpoint of target cloud	https://ovirtnode/ovirt-engine/api
Username	Username to access the target cloud	[email protected]
Password	Password to access the target cloud	password
Discovery level	Configures the level on which replication agents discover machines	Cluster
On replication start	Whether to change source machine power state on replication start. Details.	No action
On replication finish	Whether to change source machine power state on replication finish. Details.	No action

The Discovery level field configures the level on which replication agents discover virtual machines. This could be the cluster, host, data center or entire cloud level. For example, setting the value to Cluster means that one (and only one) instance of replication agent has to be deployed on each cluster. Each agent instance discover machines only from the corresponding cluster. Likewise for other levels.

Step 4. Protection instructions

Step 4 displays Download agent and the deployment instructions:

1. Select Download agent to download the RAW file, and deploy one or more VMs on oVirt, according to chosen discovery level.

2. Start deployed VMs (agents) to protect machines.

Machines will be detected and shown in a target group in a few minutes after agent deployment and launch. The detected machines appear in the selected group with the Unprotected status. Select the machines and choose Actions → Start Protection.

Alternatively, select Actions → Download Replication Agent on Manage Clouds to get the agent's file.

oVirt Agent configuration

Use the shared external-agent TUI configuration. oVirt agent logs are stored in /var/log/orca/orca.log.

OpenStack Agent

OpenStack Agent is an external replication agent type that is deployed as a separate instance in the source environment.

Requirements

Note

Replication Agent for OpenStack requires superuser permissions for installation.

• Hardware:

  • Memory: at least 4 GB RAM

  • CPU: at least 2 vCPUs

  • Disk space: at least 10 GB required. Disk will be downloaded with fix size = 10Gb. It is whole disk of new deployed VM.

• Ports for correct agent work:

  • DR host - tcp/443

  • Send logs to the Acura cluster - udp/12201

An agent instance must be deployed on each project that has machines intended for replication.

Installation

The first two steps that must be completed to successfully install the OpenStack replication agent are described in the section Machines replication and distribution between groups.

Step 3. Agent settings

Select the machine group that will contain the protected machines, configure the OpenStack cloud access settings, and select Next. To create a group first, see Create a machine group.

If the platform is not listed, select New OpenStack instead of Registered OpenStack, complete all fields, and select Next. Use the ? icons for field-specific help. The wizard validates the values before opening Step 4.

Table 5: New OpenStack fields description.

Field	Description	Example
Cloud name	The name of the cloud which will be shown in the UI. The name must be unique.	openstack_source
Keystone API endpoint	Keystone authentification URL	http://ovh-os.example.com:5000/v3
Auth type	Select the Keystone authentication type	Password
User domain	User domain name to access the target cloud	default
Username	Username to access the target cloud	username
Password	Password to access the target cloud	password
Target project domain	Target project domain where failback workloads will be spun up	default
Target project ID	Target project ID where failback workloads will be spun up	28bfd…
Hystax Service Network	Network which will be used for Hystax failback machines	internal
Floating IP Network	External network which will be used to attach Floating IPs to failback machines	provider
Source storage type	Choose source storage type for external agents. Not all agent types require or use storage credentials. Note: OpenStack Replication Agent requires Bitmap Manager for incremental replication when source storage is not provided. With Ceph storage, Bitmap Manager is required for incremental replication of ephemeral boot volumes.	Not provided
Use Bitmap Manager	Select to enable Bitmap Manager support. Bitmap Manager is an external service deployed on hypervisor nodes to track disk changes of replicated machines. Due to QEMU limitations, changes are only tracked when machines are powered on.	checked
Bitmap Manager URLs	A newline-separated list of Bitmap Manager URLs. The replication agent will query them all to search for replicated VMs.	http://bitmap-manager.example.com
Bitmap Manager token	Bitmap Manager authentication token.	token
On replication start	Whether to change source machine power state on replication start. Details.	No action
On replication finish	Whether to change source machine power state on replication finish. Details.	No action
Discovery level	Select the level on which replication agents discover machines. For example, Availability zone means one replication agent instance per availability zone in the project.	Entire project

Step 4. Protection instructions

Step 4 displays Download agent and the deployment instructions:

1. Select Download agent to download the RAW file, and deploy one or more VMs on OpenStack, according to chosen discovery level.

2. Start deployed VMs (agents) to protect machines.

Machines will be detected and shown in a target group in a few minutes after agent deployment and launch. The detected machines appear in the selected group with the Unprotected status. Select the machines and choose Actions → Start Protection.

Alternatively, select Actions → Download Replication Agent on Manage Clouds to get the agent's file.

OpenStack Agent configuration

Use the shared external-agent TUI configuration. OpenStack agent logs are stored in /var/log/orca/orca.log.

Linux Agent

Requirements

• Replication Agent for Linux requires sudo permissions for installation

• Hardware:

  • Memory: 500 MB RAM

  • Disk space: 100 MB required for product installation and not less than 15% free space of the disk size for snapshots creation

• Ports needed for the correct work:

  • Send data to Acura host - tcp/443

  • Send logs to Acura - udp/12201

Limitation

Installation and operation on systems with Secure Boot are not supported.

Installation

The first two steps that must be completed to successfully install the Linux replication agent are described in the section Machines replication and distribution between groups.

Step 3. Agent settings

Acura offers a choice between replication agent builds that depends on customer's preferences and the actual use case.

Select the previously created machine group, the target distribution (Debian/Ubuntu or CentOS/RHEL), and the snapshot driver deployment type. Then select Next.

Select Show supported kernels in Step 3 of the Download agents wizard to view the full list of natively included versions.

Pre-built and DKMS Agent types

There are two driver deployment types to choose from: pre-built and DKMS.

• The pre-built driver package (available both for Debian/Ubuntu and CentOS/RHEL distributions) requires no additional dependencies, but the supported Linux kernel list is limited. Use pre-built package for machines that are infrequently updated or have no Internet connection to install DKMS and other dependencies.

• The DKMS driver package will build the driver on installation. It has a broad Linux kernel support and will rebuild the driver on kernel updates. It requires DKMS, build tools and kernel headers to be installed on the machine. Use DKMS package for machines that are frequently updated or not supported by the pre-built package.

Select Next. Depending on the type of driver deployment you choose, the protection instructions on step four vary:

• Protection instructions for an agent with a pre-built package.

• Protection instructions for an agent with DKMS.

Step 4. Protection instructions

On the fourth step there is a Download agent button and protection instructions. An agent with a pre-built package and DKMS has different protection instructions. Follow the instructions for the selected package type:

Protection instructions for an agent with a pre-built package

1. Download a deb/rpm agent installer package file using the link above or copy the command below to run it in a terminal.

   curl -k -o hlragent.deb 'automatically generated link'

2. Copy the package to the Linux machine you would like to protect. Use deb package for Ubuntu/Debian machines and rpm package for RHEL/CentOS machines.

3. Install the agent using the following commands (superuser privileges required):

   • for Ubuntu/Debian machines: dpkg -i hlragent.deb
   • for RHEL/CentOS machines: rpm -i hlragent.rpm

The machine appears in the selected group with the Unprotected status. Select it and choose Actions → Start Protection.

Protection instructions for an agent with DKMS

1. Install DKMS, build requirements and Linux header files for current kernel and future updates on the machine (superuser privileges required):

   • for Ubuntu machines: apt-get update && apt-get install dkms perl make gcc libelf-dev linux-headers-$(uname -r) linux-headers-generic
   • for Debian machines: apt-get update && apt-get install dkms perl make gcc libelf-dev linux-headers-$(uname -r) linux-headers-amd64
   • for RHEL/CentOS machines: add EPEL repository and run yum install dkms perl make gcc elfutils-libelf-devel kernel-devel-$(uname -r) kernel-devel

2. Download a deb/rpm agent installer package file using the link above or copy the command below to run it in a terminal.

   curl -k -o hlragent.deb 'automatically generated link'

3. Copy the package to the Linux machine you would like to protect. Use deb package for Ubuntu/Debian machines and rpm package for RHEL/CentOS machines.

4. Install the agent using the following commands (superuser privileges required):

   • for Ubuntu/Debian machines: dpkg -i hlragent.deb
   • for RHEL/CentOS machines: rpm -i hlragent.rpm

The machine appears in the selected group with the Unprotected status. Select it and choose Actions → Start Protection.

Windows Agent

Requirements

• Replication Agent for Windows requires administrative permissions for installation

• Hardware:

  • Memory: 2 GB RAM

  • CPU: x64 processor

  • Disk space: 100 MB required for product installation and not less than 15% free space of the disk size for VSS snapshots creation

• Software:

  • Microsoft .NET Framework 4.0

Warning

Windows replication agent must be installed by the System Administrator or using Administrator privileges. Otherwise, it will not have enough permissions to use API or create snapshots and the replication will fail.

Installation

The first two steps that must be completed to successfully install the Windows replication agent are described in the section Machines replication and distribution between groups.

Step 3. Agent settings

Select the machine group that will contain the protected machine, and select Next.

To create a group first, see Create a machine group.

Step 4. Protection instructions

Step 4 displays Download agent and the installation instructions:

1. Download the zipped agent installer package using the Download agent button.
2. Copy the archive to a Windows machine that is intended for replication.
3. Unzip the archive and run hwragent.msi to install replication services.

The machine will be registered and shown in a target group in a few minutes after the agent installation. By default, the discovered machine will have the Discovered status. To start replication, select the machine and use Actions → Start Replication.

PostgreSQL backup agent

Requirements

• PostgreSQL backup agent is available only for Storage-type target clouds.

• Supported OS versions:

  • RHEL 9+

  • Ubuntu 21.10+

Installation

The first two steps that must be completed to successfully install the PostgreSQL backup agent are described in the section Machines replication and distribution between groups.

Step 3. Agent settings

Select the previously created machine group and the target distribution (Debian/Ubuntu or CentOS/RHEL). Then select Next.

Step 4. Protection instructions

Step 4 displays Download agent and installation instructions generated for the selected settings. Use the on-screen commands and values exactly as shown. After installation completes and the agent connects, the detected machine appears in the selected machine group.

External-agent TUI configuration

VMware, oVirt, and OpenStack replication agents use this TUI to change network settings, IP addresses, and configuration files. The main window displays the agent ID in the format AGENT_ID(CLOUD_ID)@CUSTOMER_ID, status information, and logs. It provides To console and Agent options.

Agent logs are stored at the following paths:

• VMware agent: /var/log/cabrio/cabrio.log
• oVirt agent: /var/log/orca/orca.log
• OpenStack agent: /var/log/orca/orca.log

Select To console to configure the agent from the command line, or select Agent options to use the recommended menu.

Note

Use Agent options for the recommended configuration workflow.

You can also access the TUI through SSH. Use username user and password px%96jA-8zX{ to log in, or use the SSH key configured in the cloud. Select Exit console to close the SSH session.

Console

Select To console to open the console. Use the displayed command list to configure parameters. Enter exit to return to the TUI.

To edit the machine hostname, run sudo vi /etc/hostname or sudo nano /etc/hostname.

Note

Restart the machine to apply the changes.

Agent options

Note

Select Agent options to open the menu.

Warning

Changes made here may automatically restart agent services, interrupting all running tasks. Configure the agent when it's idle.

1. Restart agent services restarts the services after confirmation. Confirm the restart in the main-window logs.
2. Enable/disable debug logging adds or removes debug information from the log. By default debug information is disabled.
3. Edit network opens the NetworkManager TUI.

1. Edit hosts opens GNU nano to edit /etc/hosts. Use it to specify DNS names manually.

1. Edit controller addresses opens the address menu for updating IP addresses and ports.

Select Update all addresses at once to update the controller address:

Select Update individual addresses to configure the REST API, receiver, report, stash, or Logstash address separately:

1. Edit agent common config opens GNU nano to edit the common configuration file. Update certificates, users, agent IDs, or controller addresses as needed.

Warning

Do not update role and cloud_type fields.

1. Edit agent service config opens GNU nano to edit the service configuration file.

Note

File content depends on the agent type.

Configure replication settings

This action is available for machines that have already been detected.

Once machines have appeared in the list and replication has been enabled for the specified machines, it is necessary to set replication schedules and snapshot storage policies. This is described in the sections Edit replication schedule and Edit retention settings.

For a Storage cloud type set backup and recovery location parameters in the Edit machine settings menu.

Create the initial consistent snapshot

A snapshot is a point-in-time representation of a protected machine. The initial snapshot serves as the baseline for all subsequent replication operations. After protection is started, the solution creates a consistent snapshot of the machine and transfers the data to the target storage. Once the initial snapshot is completed, subsequent replication cycles transfer only the detected changes, unless otherwise specified in the replication settings. This approach reduces network traffic and storage consumption while maintaining up-to-date recovery points.

Go to the Machines tab, select all machines in the group using the check box, and then select Bulk Actions → Start Protection.

Retrieve machines snapshots and detecting changes

Depending on the initial virtualization platform, it is possible to obtain consistent snapshots of machines in one of the following ways:

• in case of replicating machines on VMware, getting snapshots of Windows and Linux machines is done with standard tools of VMware (VMware CBT API), leveraging a chain of internal API calls and snapshots as well as the means of the machine's OS itself.
• in the case of replicating machines from oVirt, getting snapshots of the machines is done using standard oVirt tools. oVirt takes a snapshot at the start of replication, at the end of replication it deletes the snapshot from the previous replication, leaving a new one, an incremental replica is taken based on the definition of new blocks between two snapshots. In this regard, it is not recommended to delete the latest snapshot of the source machines. The oVirt API is used to create snapshots, and the oVirt-imageio API is used to read data.
• in the case of replicating machines from OpenStack, getting instance images with volume snapshots on each disk is done using OpenStack API. Old images (agent-created only), except the latest one from successful replication, will be removed. Reading and getting changed areas will be performed by using the Ceph RBD client for each volume snapshot. In this regard, it is not recommended to delete related images and the latest agent-created volume snapshot of the source machines to continue the replication process as incremental restore points.
• for Windows machines, consistent snapshots are acquired by means of Volume Shadow Storage (VSS) calls and tracking changes between two Application-consistent snapshots
• for Linux machines, consistent snapshots are acquired by implementation of the VSS snapshot analog for block devices and sending changes to the snapshot store.

Warning

In case of a VMware infrastructure replication, VMware Tools must be installed on your virtual machines.

Transfer deltas to storage and data deduplication

Deltas are the accumulated changes since the last replication.

Sending changes to the DC is done using the customer agent secure HTTPS protocol. This constitutes the deduplication process - the customer application only sends data that is not yet stored for the given customer allowing to significantly save the amount of transferred items, reduce load on the network and, at times, accelerate replication of machines.

In addition to WAN deduplication, optimization also includes compression of the network traffic.