Cloudflare One Migrations

14 skills · 67 min

1. Identify the source stack: Zscaler ZIA, Zscaler ZPA, Palo Alto NGFW/Prisma/GlobalProtect, legacy VPN/SWG/SD-WAN, or other.
2. Request exports and logs before mapping. Prefer structured exports over screenshots or prose summaries.
3. Build an inventory: identities, groups, apps, destinations, connectors/tunnels, DNS/URL/firewall/DLP/TLS policies, objects/lists, locations/sites, exceptions, hit counts, and compliance logging.
4. Produce a mapping plan: source object, Cloudflare One target resource, confidence, prerequisites, unsupported/partial mappings, and manual decisions.
5. Create dependencies first: identity/SCIM (opens in a new tab), connectors/on-ramps, routes/DNS, lists/objects, TLS bypasses, Access apps/policies, Gateway policies, DLP/CASB, logging.
6. Stage safely: use a migration prefix, create disabled/audit-mode rules by default, pilot with small groups/sites, compare logs, then expand rollout.
7. Account for every source rule. Each rule must map to a Cloudflare object or an explicit Not Migrated row with reason and security impact.

Exports To Ask For

• ZIA: URL filtering, firewall filtering, SSL inspection, DLP, custom URL categories, IP groups, network services/service groups, users/groups/departments, locations, GRE tunnels, and static IPs.
• ZPA: app segments, segment groups, server groups, app connectors/connector groups, access policies, IdP/group mapping, private DNS domains, ports, and protocols.
• Palo Alto/Prisma: security/NAT/decryption rules, address/service objects and groups, URL categories, HIP profiles, GlobalProtect config, Prisma Access remote network/service connection config, zones, tags, logs, and hit counts.

Mapping Heuristics

• ZIA/SWG policies usually map to Gateway traffic policies (opens in a new tab) and Gateway lists.
• ZPA private app access usually maps to Access application types (opens in a new tab), Cloudflare Tunnel (opens in a new tab), private network routing/DNS, and Access policies (opens in a new tab).
• Palo Alto rules map only after understanding traffic direction, zones, objects, users, apps, decryption, and hit counts. Do not flatten zones blindly into lists.
• Legacy VPN replacement is usually Access + Cloudflare One Client / WARP + Tunnel or Mesh for app access. Use Cloudflare WAN (opens in a new tab) only when site-to-site traffic is required; use the Network VPN migration design guide (opens in a new tab) and Replace your VPN (opens in a new tab) docs for current patterns.

Migration Assessment Prompts

• Source coverage: which products are in scope, which exports are available, and whether screenshots/prose summaries are hiding missing object files.
• Rule volume and hit data: counts by rule type, disabled/stale rules, no-hit rules, high-hit rules, and business-critical exceptions.
• Object dependencies: address objects, service objects, groups, custom categories, network services, app IDs, zones, tags, connectors, and server groups.
• Identity readiness: IdP, SCIM/group sync, group-name normalization, individual-user rules, local groups, service accounts, and contractor identities.
• TLS/DLP readiness: source decryption rules, certificate-pinned bypasses, DLP (opens in a new tab) engines/profiles, custom regex, exact-match data, and payload logging expectations.
• Connectivity readiness: source tunnels/connectors, private DNS, Split Tunnels (opens in a new tab) or bypass behavior, source IP preservation, egress IP (opens in a new tab) allowlists, and site-to-site requirements.
• Rollout readiness: pilot groups/sites, parallel-run period, rollback owner, source-stack decommission criteria, and monitoring/log comparison plan.

Source-Specific Traps

Zscaler ZIA / SWG

• Custom URL categories often split into separate IP, domain, and URL lists. Count the generated lists, not just source categories.
• ZIA locations with IPs are useful as source IP lists; they are not automatically Gateway DNS locations (opens in a new tab) for DNS policy scoping.
• GRE tunnel source IPs can inform policy conditions, but the transport migration is a separate WARP Connector or Cloudflare WAN workstream.
• CAUTION/warn behavior has no exact Gateway equivalent. Treat it as an explicit customer decision, not a silent allow/block choice.
• DLP engines and custom regex usually require manual Cloudflare DLP profile recreation. Placeholder policies must not be enabled as if DLP is complete.
• Network application groups and unsupported protocols are partial mappings. Review them before enablement.
• If SCIM is unavailable, identity-scoped source rules become overly broad unless you add an enforceable alternative such as user/email lists. Check Gateway identity selectors (opens in a new tab) before creating those rules.

Zscaler ZPA / Private Access

• ZPA app segments, server groups, and connector groups do not map 1:1. Cloudflare separates Access apps, tunnel routes, DNS, and policies.
• Creating tunnels through the API does not complete connector deployment. Plan cloudflared installation, authentication, and origin reachability separately.
• Create one Cloudflare Tunnel per ZPA connector group regardless of connector runtime status (AUTHENTICATED, DISCONNECTED, or disabled). Status is operational, not architectural. Tag disconnected or legacy groups in the tunnel description and let the customer decide what to decommission after validation.
• Each ZPA connector instance within a group maps to one cloudflared replica running against that tunnel’s token. Match replica count to connector instance count per group to preserve the same topology. A single tunnel token supports multiple simultaneous cloudflared processes. Recommend installing replicas within the same data center but on different hosts or subnets.
• For each connector group, identify all server groups linked to it and all app segments assigned to those server groups. IP addresses and CIDRs in those app segments become CIDR routes on the corresponding tunnel; domain names become hostname routes on the same tunnel. Prefer one CIDR route per subnet over per-host /32 routes where a broad subnet covers all app segment IPs.
• ZPA bypass means split-tunnel bypass in Cloudflare, not an Access bypass decision. Bypass rules map to WARP Split Tunnel (opens in a new tab) exclude entries. This is a manual configuration step with no API automation - the customer must add bypassed domains and IPs to the device profile split tunnel exclude list through the dashboard.
• Agentless/browser apps may become separate public-hostname Access apps per domain. WARP private apps remain private-destination apps.
• The default Cloudflare Access application destination limit is 5 hostnames per app. For ZPA migrations with large app segments, contact the Cloudflare account team to request an increase (up to 50) before implementation. Confirm the limit is active on the account before creating apps - without it, large segments must be split into multiple apps with identical policies, significantly increasing object count.
• IP-anchored apps require an explicit egress decision before migration: preserve source IP through customer egress, use Cloudflare dedicated egress (opens in a new tab) where available, or accept that the target service must be updated to allow new source IPs. This is a customer decision that blocks implementation if unresolved.
• Resolver policies can be account-wide. Be careful with overlapping private DNS namespaces across sites or virtual networks; retrieve resolver policy (opens in a new tab) docs before making DNS changes.
• Each ZPA access policy rule maps to a Cloudflare reusable Access policy. Create all reusable policies before attaching them to Access apps. In default-deny Gateway Network environments, additionally create a Network allow rule with selector “Self-hosted Access App with Private Address is Present” (wirefilter: any(access.private_app[*] in {"*"})) at higher precedence than any broad L4 block rules - without it, Gateway blocks private app traffic before Access policy evaluation occurs.
• In combined ZIA and ZPA migrations, Gateway Network rules can accidentally block Access private-app traffic. The Gateway Network allow rule above is the fix - place it at higher precedence (lower number) than ZIA-migrated block rules. Add and validate this rule before enabling broad L4 blocks.

Palo Alto / Prisma / NGFW

• One Palo Alto rule can produce multiple Cloudflare resources. Preserve rule intent, not rule count.
• App-ID, URL category, zone, HIP, schedule, and decryption behavior rarely translate exactly. Mark partial mappings rather than forcing false equivalence.
• Export address/service objects and groups with rules. Missing object exports cause silent-looking drops unless explicitly detected.
• Broad any destination/service rules and very broad CIDRs require manual review. Do not auto-create broad catchalls.
• HIP/device checks require Cloudflare device posture (opens in a new tab) integrations before enforcement.

Gotchas

• Source exports often split references across files. Resolve IDs against object, service, and group files before declaring a rule unmappable.
• Individual users, local groups, departments, and dynamic application IDs often need identity normalization. SCIM/group sync is the gating prerequisite for group selectors.
• Zscaler caution/warn behavior, Palo Alto App-ID behavior, and TLS/decryption exceptions may not have exact equivalents. Flag them as decision points instead of forcing a 1:1 mapping.
• Preserve source rule order and hit counts where available. Disable or delete stale/no-hit rules only with user approval.
• Never create broad allow-all catchalls to preserve connectivity unless explicitly requested and time-limited.

Validation Gates

• After each migration stage, compare Cloudflare object counts against parsed source counts. Stop on mismatches.
• Review every unsupported, partial, unmapped, needs_identity, needs_posture, and manual_review item before enabling policies.
• Validate group matching with real pilot users after SCIM sync and re-authentication.
• Test TLS inspection and Do Not Inspect behavior before enabling HTTP/DLP blocks broadly.
• Keep rollback paths explicit: disable migrated rules by prefix, restore source routing, or revert the pilot group/site.
• Before declaring done, produce a source-rule accounting table: migrated object, partial mapping, not migrated reason, security impact, and owner for each manual action.

Assessment Template