Rack clean-up and cable management
Making racks that got out of hand workable again: power, labelling and cable management — step by step, without taking anything down.
Most racks did not get messy all at once. It is years of rushed work: a cable that had to go in quickly, a label nobody updated, a plug that never quite seated. At some point nobody dares touch anything any more.
Working in a rack with no order
I do not take a running rack down to make it tidy. It gets cleaned up step by step, alongside other work, and after each step everything still runs. That takes longer than unplugging it all and starting over — and it is the only version you can afford in production.
Power that is not properly seated
Plugs sitting half in the PDU cause no fault until someone walks past. I hunt those down and seat them properly, with brackets or clips where possible. You get an overview of what was found — it usually says something about how the rack is being used.
Labelling copper and fibre
Labels that are unreadable, out of date or present on only one end get replaced with a scheme you can actually read: both ends, the same convention, and recorded in the documentation. Where conventions already exist I adopt them — this is not the moment to make everyone learn a new system.
Cable management in dense racks
Even in a densely populated rack, cabling can be laid so a server can come out without its neighbours following: cut to length, through the guides, copper and fibre separated, with bend radii that do not damage fibre.
Other services
Rack build and installation
Server racks from zero to fully operational: delivery, assembly, power, cabling and labelling — every stage handled in house.
More on thisServer builds and hardware
Building servers from scratch, assembling custom platforms and getting faulty hardware running again — including socket damage and RMA handling.
More on thisReceiving, records and shipping
Taking in equipment with a photo report, building shipment-ready pallets, and keeping a tight record of materials and hours worked.
More on thisLinux server management
Updates, configuration and day-to-day care of your servers — so you stop having to think about them.
More on thisMonitoring and alerting
Knowing something broke before your customer calls — and only being woken when it actually matters.
More on thisMigrations and upgrades
Moving to new hardware, another provider or a newer OS release — with a way back.
More on thisSecurity and hardening
Closing the basics properly: access, updates, firewall and logging. No expensive products, just measures that work.
More on thisBackup and recovery
A backup only counts once the restore has been tested. That is what I design for.
More on thisIs your infrastructure running the way it should?
Half an hour on the phone costs nothing and usually surfaces a few concrete improvements already.