distinct circumstances call for distinct restoration procedures, which can vary depending on the individuals involved, the breadth and extent of the restore, and the goal. These are a few instances of these situations:
- the unintentional destruction of emails due to human mistake, such as emptying the Trash can folder
- a file system issue that might result in reading and writing failures and render some portions of an account unavailable
- administrative mistakes, including those that occur during the undelete procedure
- Due to the use of false or stolen credentials, an account has been hacked.
- The enforcement authorities require access to an account’s whole history since they are looking into it.
In any of these scenarios, the data in a mailbox can be restored. Restore solutions fall into two categories based on the destination of the recovered data: recovery on the same server—or same infrastructure—and recovery on a different infrastructure.
reinstall the same infrastructure
These methods are designed to be applied when only a portion of an account on the same server as the origin server has to be restored. Single Item Restore, Restore on New Account, Time-range Undelete, and Restore Account all come under this category.
various infrastructure restoration
Use of the External restoration approach is a possibility when the restoration procedure is not practicable or doable on the same infrastructure as the original.
It is significant to note that objects in Carbonio Backup are available up until the point at which they are marked as deleted and removed from a mailbox as part of a Backup Purge process.
Lastly, all restoration techniques:
- may only be carried out using the CLI, with the exception of Restore Account
- retrieve things at a specific time (or interval), which means that they are also recovered in their status at that time.
- retrieving a file from a different folder than the original
- At the start and completion of the restore operation, you should always send the administrator and the procedure’s initiator an email.
If the message has been deleted or is otherwise no longer accessible, such as after being taken out of the trash, it may still be found by looking through the mailbox.log log file and looking for one of the message’s other meta-data, such as the time it was deleted, the sender or recipient, the content, etc. However, you can only use this approach if you have administrative access, so if you don’t, you should ask your admin for help.
One-Item Restoration
One of the restore modes offered by Carbonio Backup is the Single object Restore, which enables the restoration of a single object at a time while restoring its status even if it was deleted.
Single Item Restore retrieves the appropriate item from the backup using the itemID and restores it to the owner’s account. This method may be used to repair any kind of object.
A Single Item Restore is being run.
Use the command to begin an item restore procedure.
On New Account Restore
You can restore the contents of a mailbox to as they were at a certain point in time using the Restore on New Account and Restore Account methods. The two differ in the account state and the manner in which the mailbox will be restored: When an account is still active, it can be restored into a brand-new account with a brand-new accountID. This is feasible if the account was deleted and can be restored with the same accountID (Account Restore).
It is feasible to retrieve one or more deleted things from a user’s account without really rolling back their whole inbox because the source account is left untouched. You have the option to conduct this type of restoration with the newly established account hidden from the GAL as a security precaution.
This process is helpful in a number of situations, including when a whole account has been erased or is no longer functional due to a human error (such as an incorrect delete/purge operation conducted by the user or system administrator) or an external issue (hardware or storage failure).
A new account, referred to as the Destination Account, is established when a Restore on New Account operation begins. All of the things present in the source account at the chosen period, including the folder hierarchy and all of the user’s data, are rebuilt in the destination account.
Restoring a new account after a backup
Use the do Restore On New Account command in the CLI to begin a Restore on New Account.
Undelete Time-range Time-range Undelete, also known as Undelete Restore, is a Restore Mode that enables an administrator to retrieve all things that were deleted from the Trash folder within a specified time frame from a mailbox and restore their status to the point at which it was last changed.
The Carbonio Backup engine examines the backup datastore for things marked as DELETED in the given time period during a Time-range Undelete, and then restores them in the source folder of the mailbox. All objects that have been restored will be identified by the string undelete_DD_MM_YY, where DD_MM_YY is the date on which the undelete was performed.
The undeleted objects’ structure is kept, and each subdirectory is restored using the most recent attribute that was accessible during the time frame being restored. In the absence of any data, the folder will be titled unknown_XX.
A Time-range Undelete is being executed
Use the carbonio backup do Undelete command to begin a Time-range Undelete process.
Outside Restoration
The ability to import backups created on a separate infrastructure is provided by the External Restore feature, which is important for sophisticated activities like disaster recovery or putting up a test environment that is similar to the production environment. The source server and destination server cannot be the same for any other approach; this makes it the sole one.
One intriguing and practical feature of External Restore is that in addition to restoring data, it also restores all of an account’s shares.
The External Restore replicates data, information, and configuration to a new server from the Backup Path on the source server. The process is explained below and broken into three Phases. It comprises of a workflow with a number of components.
An example of a situation where External Restore comes in handy is when you need to move a server from your infrastructure in Rome to your infrastructure in Milan. The fundamental access requirement is that in order to perform an external restoration on your Milan infrastructure from the Milan server (the destination), you must have access to the Backup Path on the Rome server (the source).
Do not use Domain Provisioning
Although the External Restore is typically used on the entire infrastructure, it can also be applied to single or multiple accounts; in this case, only the data and metadata that belong to those accounts will be restored, not domain-level customizations (such as COS, GAL, quota, and so on). The skip_domain_provisioning argument may be used for this operation, as shown in the example that follows, which only restores the accounts john and alice in domain example.com:
When utilising the skip_domain_provisioning option, the process that is described below does not apply; just the Restore all Accounts’ Attributes step will be carried out in Phase 1 because no domain configuration will be affected.
Container Restore
Think about creating a folder named Inbox/Zextras (which is also its Backup Path) and subsequently deleting some messages from there that were in a backup. These messages are restored together with any other messages that may already be present in the Inbox/Zextras folder when an External Restore is performed. In other words, since the restored folder and an existent folder both share the same Backup Path, the restored messages are saved there.
What occurs is as follows in greater detail:
regional file
The backup folder id will be mapped to the existing folder id if a folder with the same path has previously been established by a filter. Additionally, everything that was in the original folder will be returned to its original location.
external mail box
The mountpoint will be restored if a filter had created a folder with that path. The filter also updates the filter to write to the restored mountpoint. All things in the folder that was produced by the filter are also relocated to the target of the mountpoint.
Before You Begin
It is expected that you have already set up a fresh vanilla infrastructure, or, more specifically, a fresh instance of Zextras that has only undergone a regular installation.
Determining a Backup Path on the new infrastructure is really the first thing to do, unless you wish to use the default one (/opt/zextras/backup/zextras), then initialising the Backup.
Additionally, you may put the following recommendations into practise to lessen the total overhead and stress on the server during the External Restore.
- Disable the RealTime Scanner if Carbonio Backup has already been initialised on the destination server to optimise memory utilisation and I/O performance.
- Advanced users can modify or deactivate the RedoLog for the duration of the import to decrease I/O overhead and the amount of disc space required for the migration.
- It is feasible to enable compression on your current main volume before beginning the import to consume less disc space overall. It is feasible to establish a new, uncompressed primary disc, set it to Current, and convert the previous one to Secondary if you do not want to utilise a compressed primary drive after migration. The powerstore component is required for this operation.
- Check out the section on multithreading to speed up the restore if you want to utilise the CLI.
Making use of an external restore
Use the carbonio_backup_doExternalRestore command to begin an External Restore procedure.
Using multithreading to expedite the restore
You may restore several accounts simultaneously with the concurrent_accounts argument, considerably accelerating the restoration process. This functionality is only accessible through CLI.
After the Restore: Message Deduplication After an External Restore, it is strongly advised to do a volume-wide deduplication using the Zextras Powerstore component because the native deduplication mechanism might not work well when importing accounts consecutively.