Service · IT infrastructure

Performance optimisation

When everything starts to stutter, the reflex is to order a bigger server or a larger Azure VM. That rarely solves it: the constraint sits in the storage, the database settings or an overloaded switch port, leaving a new CPU idling exactly like its predecessor.

Send an enquiry See the scope

Measurement

instead of guesswork

Configuration

before purchases

Before and after

documented

The cause

not the symptom

What the service covers

We work from data. Until you know which resource is maxed out, every purchase is a gamble.

Talk it through with us

Live measurement

CPU, memory, disk queue and traffic are logged during the hours your office is busiest, using our monitoring agent.

Servers

Too few resources, poor configuration, unnecessary background jobs or antivirus scanning the database files.

Databases

Missing indexes, maintenance jobs running at midday and locking when many users work at once.

Network

Port errors, mismatched speed settings and saturated links between sites or to the data centre.

Cloud and SaaS

Undersized Azure VMs, OneDrive sync eating bandwidth and latency to services hosted outside the EU.

Follow-up measurement

Fresh readings, so the improvement is backed by figures rather than claims.

How we deliver it

The analysis usually takes a working week. How long fixes take depends on what we find, and often it is only hours.

Complaints

What is slow, for whom and when. "Everything is slow" is a starting point, not a diagnosis.

Measurement

We log during peak hours and set the numbers against the complaints.

Fixes

Free options first: settings, exclusions and moving heavy jobs to the night.

Verification

New readings compared with the baseline, presented in a short Teams meeting.

A database server whose antivirus has no exclusions turns up in almost every engagement. Scanning every database file while it is in use makes everything several times slower. Configuring the exclusions recommended by the software vendor is a thirty-minute job, while users have usually put up with the slowness for months.

Common questions

Occasionally that is right, but only measurement can tell. If the storage is the constraint, a faster processor changes nothing. The analysis costs a fraction of new hardware and frequently shows the investment can wait.

Around 50 % faster is a typical result from configuration changes. Bigger jumps come when we uncover a glaring mistake, such as an index that was never created or a heavy batch job scheduled for 10am without any obvious reason.

From the numbers. We measure before and after and show you the difference. The feeling of speed shifts with mood; the measurements do not.

Related services

View the whole area

Service IT infrastructure assessment A clear picture of hardware, licences, risks and next steps Service Remote desktops and VDI Virtual desktops and secure access from home

Find out what is holding you back

Describe which tasks drag and who notices. We will measure and point to the real cause.

Email
office@apply.dk

Coverage
All of Denmark, delivered remotely

I agree that Apply may handle my details as set out in the privacy policy