Veeam Partner · Egypt
A backup you have never restored from is not a backup.
Backup, replication and disaster recovery, with an offsite copy that exists, and a recovery time that is a measured number rather than a hope.
Where we have deployed it
| Programme | What was built |
|---|---|
| Enterprise Virtualisation & Offsite DR Industrial group · Egypt | Backup architecture with an offsite copy carried over a dedicated leased line, and restore testing, so recovery time is measured rather than assumed. The network design was updated to carry replication without starving production. |
| Secure Compute Cluster Licensed e-payment provider · Tier-III DC | Storage and backup architecture under payment-grade requirements. |
The question we ask first
Not “are you backed up”. Almost everyone says yes. The question is: when did you last restore, and how long did it take?
If nobody can answer that, the backup is a line item rather than a recovery plan. We test restores so the number is known before you need it.
What we deliver on Veeam
BackupDesigned around what you actually need back first, not a default schedule.
ReplicationSized so replication traffic does not starve production.
Disaster recoveryAn offsite copy that exists and has been proven to restore.
Restore testingRecovery time measured and documented, so it is a number rather than a hope.
How we actually build a Veeam estate
Backup software is not the hard part. The hard part is a design where the thing that destroys your production data cannot also reach the copy, and where somebody has proven, with a stopwatch, how long the recovery takes. Modern ransomware goes looking for the backup server first. These are the decisions that decide whether it finds anything useful.
The rule we design to, and what each part costs
Three copies of the dataProduction plus two backups. One copy is a single point of failure with extra steps.
On two different mediaDisk plus object storage, or disk plus tape. Two repositories on the same array share a fate.
One of them off siteA copy in the same building is protected against a failed disk and nothing else. Fire, flood, theft and a determined attacker take both.
One of them immutableA copy that cannot be deleted or encrypted before its retention expires, not even by a domain administrator, and not even by someone holding your backup console credentials.
Zero errors on restore verificationThe digit everyone forgets. A job that completes successfully every night proves the job ran. It does not prove the data comes back.
Making the backup survive the attack
Hardened repositoryA Linux repository with single-use credentials and immutable backup files, so the retention period is enforced by the operating system rather than by a setting somebody can switch off.
Object storage with lockOff-site copies written to object storage with an object-lock retention policy, the same principle, in someone else’s data centre.
The backup server off the domainIf the backup infrastructure authenticates against the same directory as everything else, then compromising the directory compromises the recovery. We keep it separate, with its own credentials and MFA on the console.
Least-privilege service accountsSeparate accounts for the hypervisor, the guests and the repositories, each with only what it needs. Reusing a domain administrator account across all three is the most common finding we make.
Alerting somebody readsFailed and missed jobs escalate to our engineers. A backup estate that reports into an unmonitored mailbox is discovered to be broken at the worst possible moment.
Restores, the part that is actually the product
Instant recoveryA failed virtual machine run directly from the backup file while the full restore happens in the background, so the business is working again in minutes rather than hours.
File and folder levelThe everyday case: one deleted file, restored in minutes, without touching the rest of the server.
Application-item recoveryA single mailbox, a single database table, a single directory object. Recovered without restoring the whole machine around it.
Application-aware processingQuiescing databases and mail servers properly and truncating their logs, so what is captured is a consistent state and not a crash image you discover is unusable during the restore.
Bare-metal and dissimilar hardwarePhysical servers and endpoints recovered onto whatever hardware is actually available on the day, not onto the identical model you no longer own.
Retention, replication and the second site
Backup copy jobs with long-term retentionDaily, weekly, monthly and yearly restore points held to a schedule that matches your actual obligations. Legal, contractual and audit. Rather than a default of fourteen days.
Replication for the workloads that cannot waitA standby copy of the critical machines at the second site, ready to power on, for the systems where a restore is too slow to be acceptable.
Change rate measured firstWe measure how much data actually changes before specifying the link, then shape the traffic so replication cannot starve production.
The gaps people do not know they have
| Gap | What people assume | What is actually true |
|---|---|---|
| Microsoft 365 | “It is in the cloud, Microsoft backs it up.” | Microsoft protects the service. Your data is your responsibility under the shared responsibility model. Retention windows are short and a deleted or encrypted mailbox is recoverable only for a limited period. Mail, OneDrive, SharePoint and Teams need their own backup. |
| Snapshots | “We take snapshots, so we are covered.” | A snapshot depends on the production storage it lives on. Lose the array or encrypt the host and the snapshots go with it. |
| Endpoints | “Everything important is on the server.” | It rarely is. Laptops carry work in progress that has never been saved to a share. |
| Recovery time | “We can restore overnight.” | Until it is timed, that is a guess. Restore speed is limited by the repository and the link, not by the size of the backup. |
Where we draw the line. We would rather run a restore test on the estate you already own than sell you more licences. If the test passes and the numbers meet what the business needs, we will tell you that and there is nothing to buy. If it fails, and in most first tests something does: you will know exactly where, in writing, before it matters.
Free backup and DR review, no obligation
Tell us what you are protecting and how long you could survive without it. We will tell you whether your current position matches that answer.
Book the free assessment02 35375791