Dedicated Server Provider Due Diligence: 8 Checks That Separate Real Hosts from Resellers

A large share of “dedicated server” listings on price-comparison sites are resold capacity: the seller rents servers from a bigger provider and marks them up. That is fine until you need a hardware swap at 3 a.m. and discover the reseller has no datacenter access at all. The eight checks below take about an hour of email and phone time and will filter out most of those middlemen before you sign a 12-month contract.

Start with the basics — who owns the hardware, where it lives, and what the network actually is. Our comparison table shows which providers operate their own facilities, so compare dedicated server plans side by side as you work through the checklist.

Check 1: Who owns the hardware and the facility

Ask directly: “Do you own the servers and the datacenter, or are you reselling from another company?” Owning the facility means the provider controls power, cooling, and physical access; resellers control only a billing portal. Neither is automatically bad — some resellers offer better support than the upstream host — but you need to know which layer you are buying so you know who to call during an outage, and whether that person can actually touch hardware.

Check 2: Network quality and upstream transit

A dedicated server is only as good as its network. Ask for the provider’s ASN and check which upstream transit providers they peer with; a host on a single cheap transit line will show packet loss during peak hours, while one with multiple upstreams and major IXP peering will not. Ask how many Tier-1 carriers they buy from and whether they publish a network status page. Also ask what happens during a DDoS attack: always-on mitigation keeps you online, while null-routing takes your server offline for the duration of the attack.

Check 3: Support speed and hardware replacement SLAs

Test support before you buy. Open a pre-sales ticket with a technical question and measure the response time — that is usually the best predictor of how fast a broken hard drive will be handled. Then pin down the replacement SLA in writing. The table below summarizes the answers you want:

Question	Green flag	Red flag
Hardware swap time	4 hours or less, 24/7	“Next business day”
Spare parts location	On site in the same datacenter	Shipped from a central warehouse
Proactive monitoring	Yes, with alerts before you notice	None — you must report failures
Ticket priority for outages	P1 queue, phone + ticket	Same queue as billing questions

Check 4: Contracts, renewals, and the exit plan

Dedicated contracts typically run 12–36 months, and renewal prices can jump 20–50% if you are not watching. Confirm the renewal price in writing before signing, and check the exit terms: how much notice is required, whether you can export your data as an image, and whether your IP addresses are portable or die with the contract. A provider that makes leaving painful is a provider to be suspicious of — the exit plan matters as much as the onboarding experience.

Check 5: Bandwidth metering and burst policy

Unmetered does not mean unlimited. Most providers cap sustained throughput at a fair-use threshold, then shape or bill you above it. Ask for the exact threshold in writing, and ask how burst traffic is treated: a game server that spikes during a launch weekend can blow through a metered allocation in hours, and a 1 Gbps “unmetered” port is often fair-use capped around 5–10 TB/month. If you cannot get a straight answer on the metering policy, assume the worst and budget for overage fees.

Check 6: Control panel, API, and automation access

You will reboot, reinstall, and monitor this machine for years, so ask what self-service tools are included. A usable control panel with one-click OS reinstalls, IPMI/KVM access, and usage graphs is standard at good providers. If you plan to automate, ask whether the panel has an API and whether it can trigger reboots, change boot order, and attach ISO images remotely. Providers that lock everything behind support tickets will cost you hours every time something needs changing.

Check 7: Security and compliance posture

Ask how the provider handles compromised servers: do they quarantine the machine, notify you, and preserve logs, or do they simply null-route you until you “fix it”? Find out whether the datacenter is audited (SOC 2, ISO 27001) and whether backups are stored off-site and encrypted. If you handle payment data or personal information, these answers determine whether you can legally use the provider at all.

Check 8: Pricing transparency and long-term value

Finally, total the real cost: setup fees, IP address rentals, backup storage, DDoS protection add-ons, and the renewal rate. A $120/month server with a $50 setup fee and mandatory $20/month backup add-on costs more than a $150/month server with everything included. Compare all-in pricing, not advertised pricing, and ask for the price in writing with a 30-day validity so the quote cannot change when you come back to sign.

Run these eight checks against your shortlist, then see the full specs in our provider table to compare pricing and contract terms side by side before you commit.

InterServer is one of the few providers that owns its datacenter and publishes hardware replacement terms openly — browse the current InterServer dedicated server lineup to see how their contracts stack up against this checklist.