Peak-time readings
CPU, memory, disk latency and network traffic recorded during the busiest stretches, month-end close being a classic.
Every case starts with data from the hours when people complain. Without it, any purchase is a coin toss.
CPU, memory, disk latency and network traffic recorded during the busiest stretches, month-end close being a classic.
Missing indexes, stale statistics, a reindex job firing at 10 am and locking when a dozen people post entries at once.
Spinning disks under a database, a RAID 5 array mid-rebuild, a crowded NAS holding backups on the same volume as live files.
Ports on the wrong duplex setting, overloaded Wi-Fi in an open-plan office, and branches reaching the head office database over VPN on a basic line.
Undersized Azure VMs, bottom-tier disks, OneDrive sync hogging bandwidth and services hosted far outside the EU.
WooCommerce, PrestaShop or Shoper with integrations: caching, PHP-FPM, plugin queries and a BaseLinker sync that chokes the database every quarter of an hour.
How long diagnosis takes depends on the system. Many fixes need no interruption to work.
What is slow, for whom and at what time. We ask for specifics: a named report, a particular page, an exact hour.
An agent gathers data over several working days, which we line up against the times of complaints.
Free changes come first: database settings, job schedules, antivirus exclusions, network configuration.
The same operations are timed again and the results, in seconds, are shared on a short video call.
A branch working on a head-office database over VPN will usually be slow. Accounting software fires hundreds of tiny queries every time a window opens, and each has to cross the line. No server can cure that. What works is moving the branch onto a terminal server or virtual desktop, so only screen images travel over the connection.
Yes, provided we start a few weeks ahead. We time page generation, review caching and plugin queries, then simulate traffic several times the usual level. Changes go live in a quiet period, never during promotion week.
That is exactly why we measure. With data in hand we talk to Comarch, enova365 or whichever vendor it is and show where the bottleneck sits. Sometimes it really is hardware, but just as often it is a database setting the vendor recommends that nobody ever applied.
Mostly not. Database settings and job schedules can be changed during the day. If a server needs a restart or a database has to move to another disk, we book an after-hours slot and give the team notice.
It is billed at PLN 190 per hour excl. VAT or covered by an ongoing contract. You receive an estimate of hours and a quote before we begin.
Describe one or two specific cases, such as a report that takes ten minutes. We start by measuring.
Your enquiry has reached us
You will hear back within one working day, and if you have reported an outage that is holding up work, it goes to the front of the queue.
No match for that name. Try a different spelling or pick a bigger town nearby - all our support is delivered online, so your choice has no effect on the service.