Month-end, invoicing day. The Pantheon database will not open because free space on the server had been shrinking for weeks while warning emails went unread. That is how most stoppages look: plenty of hints in advance, and on the morning itself it all feels like rotten luck. Slovenia adds its own extras, such as summer storms knocking out electricity for half a day, fibre torn up by roadworks and floods that keep a building shut for days. Apply keeps an eye on the early hints from afar and prepares you for whatever cannot be avoided. That means a backup that restores when asked, an alternative internet link, servers that power off gracefully during a blackout and a plan on paper that people can work through without panic. Our opening question never changes: what does sixty minutes of downtime cost your firm?
We begin with consequences rather than kit. As soon as it is clear which service must keep running at all costs and which can stay dark overnight, the budget more or less allocates itself.
Everything fits on a single sheet. ERP and bookkeeping, the e-invoice exchange, your online store, mail and shared drives each get two questions. How long may it be unavailable (RTO)? How much recent work could you afford to re-enter manually (RPO)? Bookings at a 60-room hotel at Bled might tolerate no gap at all, whereas old project files can wait a week.
Disks logging SMART faults, partitions beyond 90% capacity, certificates and domain registrations close to expiry, subscriptions ending, a service that crashes and restarts nightly, ageing UPS cells, a backup that quietly reported failure. On Premium someone acts on these alerts whatever the hour; Start and Business customers see them handled in office time.
Beyond the usual 3-2-1 rule, one copy is written as unchangeable to a separate store with its own logins and its own second factor. Even an attacker in full control of your server has no way to wipe it. We keep versions for long enough that an error nobody notices for weeks can still be undone.
A second carrier, by fibre or mobile; a standby replica inside the EU for the handful of systems that cannot pause; and a UPS programmed to close servers one after another during a blackout. Your own people or a nearby supplier change the failed disk or battery while an Apply engineer talks them through it on camera.
Plenty of downtime is self-inflicted by a patch. We therefore trial updates on a few machines before releasing them company-wide during a slot you have signed off. Whenever the platform supports it, a snapshot is taken first, so a sector application that refuses to start is reverted in minutes.
Four pages: who calls the shots, the restart sequence, messages to employees and suppliers, keeping customers served on paper, plus pre-written wording for the website and social channels. A section covers home log-ins for days when the office cannot be entered. The audience is the management team, not IT, and copies live outside the server room.
Nothing dramatic, all predictable. Every one has a matching countermeasure, and none of the countermeasures covers them all.
Countermeasure: track expiry dates. Right in the pre-Christmas rush the store shows a “not secure” warning and card payments hang during 3-D Secure. A heads-up a month ahead and an automatic renewal that has been proven to work make sure it never happens.
Countermeasure: alerts and a copy. Out of space, Pantheon or Minimax can no longer save, no invoice can be raised and dispatch has no delivery notes. When a drive dies completely, it is swapped under warranty by your supplier or staff, and we restore the operating system and the data on it.
Countermeasure: an unchangeable copy. Criminals go looking for your backups before anything else and lock them too. One copy kept behind different logins and a different second factor is what separates reopening the same afternoon from negotiating with extortionists.
Countermeasure: UPS, controlled shutdown, cellular backup. Groundworks on an industrial estate slice through the fibre, or a thunderstorm blacks out an entire valley all afternoon. The servers must stop without wrecking their databases and come back in the proper order when the electricity does.
Countermeasure: snapshot, then an approved window. Think of a WooCommerce extension broken by its own upgrade, or accounts software crashing after Windows Update. Holding a snapshot means a few minutes' inconvenience; lacking one can mean a lost day.
Countermeasure: keep data and logins independent of the premises. Flood water, a tremor or a fire may close the office for several days. When the sole backup was a drive sitting next to the server, recovery has nothing to work with. Copies stored in another town or country, plus cloud accounts people can use from anywhere, are the foundation for this case.
Consider a furniture exporter in the Savinja valley whose “nightly backup” has been trusted for years. When the server finally gives up, everyone learns that the job was copying database files while they were in use, so it had been storing broken copies that cannot be opened. Failure notices had gone, for two years, to the inbox of someone who had long since left. A green status line shows none of this. Only an actual restore does. So at the intervals we agree, Apply restores for real and logs the time it took.
It is. When a provider pledges, say, 99.9% uptime, that pledge is about its data centres, not about your files or settings. A user wiping a whole SharePoint library, ransomware scrambling a VM or an upgrade breaking an app all leave the provider's promise intact while your business sits idle. Protecting your data, watching your services and knowing how to get back stays with you.
Settle three matters in good time. First, keep at least one data copy far away, preferably in an EU data centre well outside your region. Second, make sure colleagues reach mail and business apps from home rather than only through the office server. Third, print a short call tree for managers showing who contacts whom when mobile networks and broadband are unreliable. Each has its own chapter in the plan.
Mainly from how short your RTO is. Coming back within the working day is affordable; coming back inside an hour means a warm standby and noticeably more money. Once the analysis is done, most companies find that only two or three systems justify serious protection. We charge €50 an hour plus VAT for the first assessment and a single restore drill, giving you an estimate of the hours up front. Customers on Start, Business or Premium get recurring drills as part of their plan.
With Premium, alerts are acted on at any time, and you receive the duty number together with your contract. Start and Business cover weekdays between 9:00 and 17:00; weekend cover is available by adding 24/7 support. Engineers always work remotely, and when hardware needs a hand they direct a member of your staff using the emergency plan.
Quarterly full recoveries for the systems you cannot live without, and one or two a year for everything else, make a reasonable starting rhythm. Big changes deserve an extra drill: a server migration, an ERP version jump or a new backup product. Results, timings and any hiccups are noted in a short log after every drill.
Generally it will. GDPR lists timely restoration of access to data among its security measures, NIS2 and ZInfV-1 expect essential and important entities to take appropriate continuity measures, and ISO 27001 addresses continuity too. A lawyer should confirm which obligations actually reach your company. Whatever the answer, an impact analysis, logs of recovery drills and a written plan are exactly the evidence auditors and large buyers request in vendor assessments.
We examine your backups and alerting, run a genuine recovery and report back, stopwatch in hand, how long each critical system needs to return.
Enquiry received
An engineer will write back no later than the following working day. Faults that leave your staff unable to work are always looked at first.
No town by that name here. Check the letters, or settle for the nearest larger place. Our engineers work only over the internet, so Kočevje and Koper get precisely the same treatment.
