Area 07 · Operations and infrastructure

Backup and recovery

Picture a textile exporter in Gaziantep. Late on a Friday its file server gets encrypted, and on Monday morning the export paperwork for the customs broker and the open orders in Logo are out of reach. The backup console has shown a green tick every night for a year, yet nobody ever timed how long bringing back an entire server would take. That gap is where we begin. Which system may stay offline for how long, and how many hours of work can you stand to redo? Only then do we choose tools, storage locations and retention rules. The job ends with a genuine restore, clock running. All of it, drills included, is carried out remotely.

3-2-1
data held three times, on two media types, once away from the office
Report
for every restore drill, dated and with the measured duration
EU · TR
your choice of data centre for the off-site copy
Recovery plan
written down while things are still calm

Targets first, tools second

For each system we agree on two numbers. RTO is the longest it may stay down; RPO is how much work you are prepared to lose in the worst case. The figures below suit many SMEs with 10 to 50 staff. Treat them as an opening position for the conversation rather than a rule.

Accounting and ERP

Four hours down, one hour lost. A single overnight copy of Logo, Mikro or Netsis is not enough on the day the VAT return is due. The SQL Server database also needs transaction log backups during working hours.

File server and SharePoint

One working day on both counts. An evening job plus shadow copies taken during the day usually covers it, so a spreadsheet someone saved over by mistake is back within minutes.

Microsoft 365 and Google Workspace

Four hours down, one hour lost. The recycle bin only protects you for a limited time. A thread deleted months ago, or the mailbox of someone who has left, normally comes back only from an independent backup of the tenant.

Online store

One hour down, a few minutes lost. When orders flow in from your WooCommerce site as well as Trendyol and Hepsiburada, and payments go through iyzico or PayTR, restoring from a nightly file takes too long. What works here is a replicated database and a standby environment prepared in advance.

Clinic and practice software

Two hours down, one hour lost. Without patient records and the appointment book, a polyclinic cannot see anyone. Because health information is a special category of personal data, the copy must be encrypted and access to it tightly defined.

Project and drawing files

One working day down, a few hours lost. In architecture and engineering offices, Autodesk and BIM models often sit on designers' laptops. The target is only realistic once those files live in a shared location that is itself backed up.

Your backups should not trust your own network. When the backup console lets people in with their everyday domain logins, whoever gets hold of an administrator credential will delete the backups first and then encrypt everything at leisure. So we give the backup platform separate accounts protected by multi-factor sign-in, and make one or more storage targets immutable, so no admin account can clear them ahead of schedule.

Our method

Four stages, all done through a protected remote connection to your servers, NAS units and cloud accounts. The first three are routine craft. The fourth, a restore with a stopwatch, is the one most providers skip, yet it is the single stage that actually demonstrates something.

01

Tracking down the data

We look at where people really work. Something outside the backup turns up almost every time: quotes on the sales manager's desktop, a sector application's database installed on a second drive, price lists the team shares from a personal Google Drive.

02

Agreeing the targets

Together we fix acceptable downtime and acceptable data loss for each system. Those numbers drive the budget; we do not bend the numbers to fit a budget.

03

Building and monitoring

Schedules, retention, encryption and the off-site copy. Alerts fire on errors and also when an expected report simply never arrives, because a job that quietly stopped is the most dangerous kind.

04

Proving the restore

On an agreed schedule we restore a sample of systems into a sandbox and note how long it took. You can show the dated report to management, an auditor or your cyber insurer.

Frequently asked questions

It beats nothing, but it has three weak spots. If disaster strikes on a Thursday, almost a week of work is gone. The drive often stays plugged in after the copy, so ransomware encrypts it too. And when that colleague goes on leave or moves on, the routine silently stops. You need not throw the drive away; we make it one piece of an automated, monitored scheme that also keeps a locked copy off-site.

Microsoft and Google keep the service running and make sure a hardware fault in their own facilities does not cost you data. Against deleted folders, items that aged out of the recycle bin, a hijacked account or malware encrypting OneDrive files, that protection only goes so far. An independent backup with retention rules you control fills the hole.

They can, as long as the attackers do not reach them first. Backups are what you fall back on when everything else has failed. Today's gangs frequently roam a network for days hunting for NAS devices and backup servers to wipe before they encrypt anything. A locked copy with an outside provider and a restore tested recently let you get running again without negotiating. When personal data is affected, KVKK may require notifying the Board within 72 hours, and you can also report the incident to USOM, Türkiye's national cyber incident response centre. Your legal adviser drafts the notification; we provide the technical findings.

We usually start from 30 days of dailies, roughly 13 weeklies and 12 monthlies. For accounting records and e-documents the retention periods under the Tax Procedure Law apply, so have your mali müşavir confirm what must be kept and for how long. There is a flip side: KVKK expects personal data not to be held longer than needed, and your retention and destruction policy covers backups as well. We turn both demands into concrete rules inside the backup software.

There are two parts. Storage is the provider's monthly fee for the volume held in the cloud, which is modest for most SMEs. Labour covers the review of what you have now and the build, charged at €55 per hour plus VAT against an estimate given up front, or as a fixed-price quote when the scope is clear. Ongoing monitoring and drills can be folded into the Start, Business or Premium plan.

The storage box or server flags the failure itself, and it usually reaches our screen before yours. We send you the correct drive model and a photo with the right bay marked. A member of your staff fits it while we guide them on a video call, or a local technician you already use does it. After that we supervise the array rebuild and verify the data, all from a distance. Apply does not come on site and does not sell hardware.

Let us test whether your backup really comes back

We review your current set-up remotely, attempt a real restore and tell you plainly whether it would carry you through an incident.

Availability
Weekdays 09:00-18:00 Turkey time (GMT+3); an answer follows by the next working day
Calls
By video, over Microsoft Teams or Google Meet

The only cookies here are the essential ones: they keep the site running and remember the city you picked. Nothing is used for advertising or tracking. See our privacy notice for more.