AllTech
01
DNS lives at the registrar, or at the old web host
Single-region nameservers, no redundancy, no monitoring, and a control panel someone's former web designer owns.
Cloudflare Partner · SASE Solutions
If your DNS goes down,
everything goes down with it.
DNS is the one dependency every other system has. Your website, your email routing, your VPN endpoint, your line-of-business portal — all of it resolves through DNS first. AllTech moves that record set onto Cloudflare's anycast network, signs it, and puts always-on DDoS mitigation in front of everything you publish.
Cloudflare Partner
Anycast DNS
Always-on mitigation
Local engineers
01 · The problem
Nobody owns your DNS.
In most networks we take over, DNS is a footnote. It was configured once, by somebody who has since left, at a registrar nobody has the login for. It works until it doesn't, and then every system fails at the same time. The recurring findings:
02
No DNSSEC
Nothing cryptographically proves your answers are yours, which leaves room for cache poisoning and resolver spoofing.
03
Records nobody can explain
Stale A records pointing at decommissioned servers, duplicate SPF records breaking mail authentication, TXT entries from a vendor trial in 2019.
04
The origin IP is public
Even behind a proxy, an exposed IP in an old DNS record or mail header lets an attacker bypass every protection you paid for and hit the server directly.
05
No DDoS plan beyond calling the ISP
A volumetric attack saturates the circuit before the firewall gets a vote, and the ISP's answer is usually to null-route you — which is indistinguishable from the outage you were trying to prevent.
06
Nobody is watching the TTLs
In an incident, a 24-hour TTL means a 24-hour recovery.
02 · What we deploy
Cloudflare DNS and DDoS, run as a managed service.
We move authoritative DNS to Cloudflare's anycast network, publish it from every edge location simultaneously, and enable mitigation in front of the records that face the public internet. Configuration is documented and version-controlled alongside the rest of your Cloudflare tenant — so the next engineer can read it.
01
Authoritative DNS
Your zone served from Cloudflare's global anycast network. Every nameserver answers from the nearest edge, so resolution stays fast and no single site failure takes the zone offline.
02
DNSSEC
Zone signing enabled and the DS record published at your registrar, so resolvers can verify answers came from you and were not altered in transit.
03
Always-on L3/L4 DDoS mitigation
Volumetric floods — SYN, UDP amplification, DNS reflection — absorbed and dropped at the edge, automatically, with no need for anyone to notice and flip a switch first.
04
L7 application-layer mitigation
HTTP floods, bot traffic, and application-targeted requests filtered with managed rulesets, rate limiting, and challenge actions rather than blanket blocks.
05
WAF & managed rules
Cloudflare's managed rulesets in front of any public web application, tuned per site instead of enabled and forgotten.
06
Zone hardening
Record audit and cleanup, SPF/DKIM/DMARC verified before cutover, sane TTLs, proxied vs. DNS-only decisions made record by record, and origin IP concealed where the record is proxied.
07
Optional: Spectrum / Magic Transit
For TCP/UDP services that aren't HTTP, or for protecting your own IP space and circuit rather than just your published records. Scoped separately — see limitations.
03 · Why anycast matters
The attack lands everywhere, so it lands nowhere.
01
Distance
Queries and requests are answered at the edge nearest the user, not backhauled to one data center you rent.
02
Capacity
An attack aimed at your domain is spread across the entire network rather than concentrated on one circuit. Absorption replaces heroics.
03
Failure domain
No single nameserver, region, or link is the thing that takes your zone down.
04 · How we engage
Three ways to start.
01
ASSESS · DNS & exposure review
Zone audit, record inventory, mail authentication check, DNSSEC status, and origin-exposure check. Written findings and a migration plan. Often the right first step, and it stands alone.
02
DEPLOY · Migration & hardening
Staged zone build, verified cutover, DNSSEC enabled, mitigation and WAF configured, documentation handed over. Fixed scope.
03
OPERATE · Managed DNS & mitigation
Record changes, rule tuning, attack response, certificate and TTL management, monthly review. When something breaks at 2 a.m., the people who built it answer.
05 · Migration
Four steps, and the risky one is the one everyone skips.
- 01
Audit
Export the existing zone, inventory every record, identify what's live and what's dead, confirm registrar access. Risk: None — read-only.
- 02
Stage
Build the zone in Cloudflare, match records exactly, decide proxied vs. DNS-only per record, lower TTLs at the current provider. Risk: None — not yet authoritative.
- 03
Verify
Compare staged zone against production record by record. Mail authentication (MX, SPF, DKIM, DMARC) checked explicitly, because this is where migrations break. Risk: This is the step that prevents the outage.
- 04
Cut over
Update nameservers at the registrar, watch propagation, then enable DNSSEC and mitigation once resolution is confirmed clean. Risk: Reversible — old zone stays in place.
Mail is the failure mode. A web record that's wrong is visible in seconds; an SPF record that's wrong is silent for a week and then your invoices are in a spam folder. We verify mail before we touch nameservers.
06 · Under attack
What actually happens when it starts.
01
Detection
Mitigation is on before the attack, not configured during it. L3/L4 floods are dropped automatically at the edge.
02
Escalation
For application-layer attacks we tighten rate limits, raise challenge posture, and — if warranted — enable Under Attack Mode for the affected hostnames.
03
Communication
You get told what's happening, what we changed, and what the user impact is. In plain language, while it's happening.
04
After
Written summary: vector, duration, what was blocked, what rules we're keeping, and what we're recommending you change.
If we run your Managed SOC, this is already in scope. If we don't, it's still covered under managed DNS — the OPERATE tier above.
07 · Who this is for
Where it makes the most difference.
01
Anyone whose revenue runs through a website or portal
E-commerce, scheduling, client logins, payment pages.
02
Multi-site organizations
One zone, one set of records, one place the truth lives — instead of DNS split across three vendors and two former employees.
03
Businesses recovering from an outage or an attack
If you've just been null-routed by your ISP, this is the conversation.
04
Anyone with a cyber-insurance questionnaire in hand
DNS integrity and DDoS mitigation show up on renewal forms now.
05
Organizations that cannot find their registrar login
Genuinely common. We fix ownership as part of the engagement.
08 · Honest limitations
What this will not do for you.
01
It does not protect your internet circuit.
Cloudflare protects what's published through it. A flood aimed at your ISP-assigned IP range still saturates your own pipe. That's what Magic Transit exists for, and it's a separate, larger scope.
02
A leaked origin IP defeats it.
If an attacker learns your origin address from an old DNS record, a mail header, or a certificate log, they can bypass the edge entirely. We lock the origin firewall to Cloudflare ranges — but this only works if we're allowed to.
03
Non-HTTP services need Spectrum.
Proxying protects HTTP/HTTPS. SSH, RDP, SMTP, game and VoIP protocols need Spectrum or Magic Transit, on higher tiers.
04
Some features are plan-gated.
Rate limiting depth, WAF ruleset control, and certificate options vary by Cloudflare tier. We tell you which tier your requirements need before you buy one.
05
DNS migration has a propagation window.
Recursive resolvers honor cached TTLs. We lower TTLs in advance to shrink it, but we can't eliminate it.
06
It is not a WAF strategy by itself.
Managed rules block known patterns. Business logic abuse, credential stuffing, and API abuse need tuning and review, not just an enabled toggle.
07
It doesn't fix an application that falls over at 200 concurrent users.
Mitigation stops attack traffic. It doesn't make an undersized origin fast.
Related: Secure Web Gateway · Managed SOC · Network & Infrastructure
09 · FAQ
Common questions
Will moving DNS to Cloudflare break our email?
Not if it's done properly. MX, SPF, DKIM, and DMARC records are migrated and verified before nameservers change. Mail records are the most common cause of a bad DNS migration, which is why verification is its own step.
Do we have to move our domain registration?
No. Cloudflare becomes your authoritative DNS while your registrar stays where it is. You update two nameserver entries. We can also consolidate registration later if you want to.
How long is the outage during migration?
There isn't one, when it's staged correctly. The Cloudflare zone is built and verified before it becomes authoritative, and the old zone stays in place as a fallback.
What counts as a DDoS attack big enough to matter?
For a small business, far less than you'd think. Attacks well under the headline-grabbing sizes are more than enough to take a single office circuit or an unprotected web server offline.
Is mitigation something we have to turn on when attacked?
L3/L4 mitigation is always on. Application-layer response has an escalation path we manage, including Under Attack Mode when the situation warrants it.
Does this replace our firewall?
No. Your firewall controls traffic at your edge. This protects what you publish to the public internet. Different jobs; you want both.
Can we keep internal DNS on our own servers?
Yes. Internal resolution stays where it is. This is about your public authoritative zone. For internal hostname routing for remote users, see our Secure Web Gateway page.
What does DNSSEC actually get us?
It lets resolvers verify that the answer for your domain genuinely came from your zone and wasn't tampered with in transit. It's a signature, not a filter — cheap to enable and increasingly expected on security questionnaires.
10 · Related services
The rest of the stack it plugs into
01
Cloudflare Zero Trust
The full Cloudflare One platform this sits inside.
02
Secure Web Gateway
Outbound filtering — this page is the inbound half.
03
Zero Trust Access (ZTNA)
Publish internal apps without exposing them to the internet at all.
04
Managed SOC
Humans reviewing what the edge flags.
05
Network & Infrastructure
The circuit, firewall, and multi-WAN failover underneath.
06
Backup & Disaster Recovery
For when the problem is the origin, not the traffic.
Start with a conversation
Find out what your DNS is actually publishing.
Send us your domain and a short note about your environment. We'll run a zone and exposure review, tell you what needs fixing, and be straight with you if what you have is already fine.
