AZ-700 · CHEAT SHEET
Microsoft Certified: Azure Network Engineer Associate
Key concepts grouped by official exam weight.
Download as PDF (free, no sign-up)1. Design and implement core networking infrastructure
25–30%Design and implement IP addressing for Azure resources
- Every virtual network you plan to connect to other networks or to on-premises needs a non-overlapping address space, and Azure reserves five addresses in each subnet (the first four and the last). You can extend or shrink a range while it still contains its subnets; in a peered network, every range change requires a peering sync.
- A virtual network belongs to a single region and subscription, and resources that connect to it must be in the same region; for another region you create another network and peer it. When you create it you define one or more non-overlapping CIDR ranges (RFC 1918, RFC 6598 or your own public ranges), never link-local, multicast, loopback or 168.63.129.16.
- Several services require subnets with a reserved name and no other resources: GatewaySubnet for virtual network gateways, AzureFirewallSubnet (and AzureFirewallManagementSubnet when the management NIC is enabled) for Azure Firewall, and AzureBastionSubnet for Azure Bastion. Application Gateway uses a dedicated subnet with a name you choose, while private and service endpoints are configured on ordinary subnets.
- Subnet delegation gives a PaaS service permission to create its resources in a specific subnet, for example Microsoft.Sql/managedInstances or Microsoft.Web/serverFarms. The owner can still apply NSGs and routes, but a delegated subnet doesn't support private endpoints, and the delegation can't be removed while resources of the service are deployed.
- Some services require a dedicated subnet that can hold only their own resources (Route Server, SQL Managed Instance, Application Gateway, gateways in GatewaySubnet), while virtual machines, network virtual appliances and private endpoints can share a subnet. In a shared subnet, NSGs affect private endpoints only when their network policy is enabled.
Design and implement name resolution
- Inside a virtual network you can use Azure-provided resolution (no configuration, scoped to that network), Azure Private DNS zones (the preferred option), Azure DNS Private Resolver, or your own DNS servers that forward to 168.63.129.16. Azure-managed reverse DNS is scoped to each network; to span several, you use an in-addr.arpa private zone with manually managed PTR records.
- The DNS servers setting of a virtual network can be Default (Azure-provided) or a list of custom servers, which are used in order and not round-robin. The DNS setting of a NIC overrides the network setting, and after any change existing VMs must renew their DHCP lease or restart to apply it.
- A public Azure DNS zone hosts the records of a domain and is published by delegating the domain to its four name servers at the registrar; subdomains are delegated with an NS record set in the parent zone. A CNAME can't exist at the apex, so alias records are used, which also follow the life cycle of public IPs or Traffic Manager profiles and prevent dangling records.
- An Azure Private DNS zone is a global resource that resolves names only for linked virtual networks, in any region and without peering. It supports split-horizon with a public zone of the same name; autoregistration creates A records only for VMs and each network can have a single registration zone, so other resources are added manually.
- Public zones are created with az network dns zone create and private zones with az network private-dns zone create; records use relative names and are grouped into record sets by name and type, with one TTL per set. A public zone is published by updating the NS records at the registrar with its name servers, and a private zone resolves only from linked virtual networks.
Design and implement VNet connectivity and routing
- Service chaining steers traffic from a spoke to an NVA or a VPN gateway in a peered network by using user-defined routes. Gateway transit is enabled on the hub side and use remote gateways on the spoke, which then can't have its own gateway. You can't route between virtual networks with a UDR whose next hop is an ExpressRoute gateway.
- Peering connects virtual networks in the same or different regions over the Microsoft backbone. It needs nonoverlapping address spaces and a peering on each network: the first stays Initiated until the reciprocal one exists. It isn't transitive, doesn't resolve names with default Azure DNS, and must be deleted before you move the virtual network.
- Azure Virtual Network Manager collects virtual networks into network groups, with static membership or dynamic membership through Azure Policy, and applies mesh or hub-and-spoke connectivity configurations to them. Direct connectivity builds a regional connected group that global mesh extends across regions. Nothing takes effect until deployed per region, and each deployment sets that region's goal state.
- User-defined routes live in a route table and override system routes. They support the VirtualAppliance, VirtualNetworkGateway (VPN gateway only), VnetLocal, Internet and None next hop types, and they can use service tags as the address prefix. An NVA used as a next hop needs IP forwarding enabled on its NIC and should sit in a different subnet to avoid loops.
- A route table is associated with subnets, never with the virtual network, and each subnet can have at most one. It can be associated only with subnets in the same location and subscription as the table, and it can't be deleted while associated. Don't associate a table with 0.0.0.0/0 to GatewaySubnet when there's a VPN gateway; after dissociation the subnet uses default routes.
Monitor networks
- Network Watcher is enabled per region, in NetworkWatcherRG. Virtual network flow logs go to a standard storage account in the same region and can be set at NIC, subnet or VNet level, with precedence in that order; traffic analytics needs a Log Analytics workspace. Connection monitor and packet capture need agents, and saving captures to storage requires key access.
- Network Watcher bundles diagnostic tools: IP flow verify shows which rule allows or denies a TCP or UDP flow, NSG diagnostics also covers ICMP and service tags, and effective security rules aggregate the NIC and subnet rules. Connection troubleshoot tests at a point in time, while Connection monitor continuously tracks latency and loss.
- Azure Monitor network insights, formerly Azure Monitor for Networks, shows topology, health, metrics and alerts for all network resources across subscriptions without prior configuration. Its connectivity view depends on Connection monitor tests, its traffic view on flow logs and traffic analytics, and the diagnostic toolkit opens the Network Watcher tools.
- DDoS Network Protection is activated by linking virtual networks to a protection plan, which works across subscriptions in the same tenant and includes rapid response and cost protection; DDoS IP Protection is enabled on each Standard public IP without a plan. To monitor, use the Under DDoS attack or not metric in alerts and the DDoSProtectionNotifications and DDoSMitigationFlowLogs categories in Log Analytics.
- Secure score groups Microsoft cloud security benchmark recommendations into controls such as Secure management ports or Restrict unauthorized network access; a control earns all its points when every resource meets all its recommendations, and preview recommendations don't count. You can remediate with Fix, prevent regressions with Enforce or Deny, and query controls with Azure Resource Graph.
2. Design, implement, and manage connectivity services
20–25%Design, implement, and manage a site-to-site VPN connection
- A VPN gateway has two instances, active-standby by default; active-active mode gives each instance a public IP and requires the on-premises device to build a tunnel to each. For the highest availability, combine it with two on-premises devices, one local network gateway and connection per device, and BGP. Zone redundancy requires an AZ SKU and zone-redundant Standard public IPs.
- New VPN gateways use the AZ SKUs (VpnGw1AZ to VpnGw5AZ), which support BGP, active-active, custom IPsec/IKE policies and zone redundancy; non-AZ SKUs are being migrated and can't be used to create gateways. NAT requires VpnGw2AZ or higher. Basic is for development only: no BGP, RADIUS, IPv6, active-active or zone redundancy.
- A site-to-site connection links a route-based VPN gateway to a local network gateway with a shared key that must match the on-premises device. In active-active mode the device needs a tunnel to each public IP, and both tunnels form one connection. Connection Mode sets who initiates IKE, and you check the status with az network vpn-connection show.
- Policy-based VPNs encrypt by prefix combinations and exist only on the Basic SKU with IKEv1; route-based VPNs use any-to-any selectors and are required for point-to-site, BGP and active-active. The type can't be converted. For on-premises policy-based devices, use a route-based gateway with a custom IPsec/IKE policy and UsePolicyBasedTrafficSelectors.
- The local network gateway represents the site and its VPN device: it stores the device's public IP or an FQDN, the on-premises prefixes Azure routes to it and, with BGP, the BGP peer IP and the on-premises ASN. Changing prefixes with Set-AzLocalNetworkGateway overwrites the previous ones, and modifying it while a connection exists can drop the tunnel.
Design, implement, and manage a point-to-site VPN connection
- Point-to-site needs a route-based gateway; new deployments use an AZ SKU, because Basic supports IKEv2 with certificates but not OpenVPN, RADIUS, IPv6 or active-active, and therefore not Microsoft Entra ID either. The number of P2S connections depends on the SKU, and an active-active gateway with P2S needs a third public IP.
- Point-to-site supports OpenVPN (TLS over TCP 443, cross-platform and the only one that works with Microsoft Entra ID), IKEv2 (IPsec over UDP 500 and 4500) and SSTP (TLS for Windows only, being retired). SSTP and IKEv2 can coexist, but SSTP and OpenVPN can't. After changing the tunnel type or authentication, generate and distribute a new VPN client configuration package.
- A P2S gateway supports three authentication types: Azure certificate, Microsoft Entra ID, and RADIUS with Active Directory, and you can combine them. Microsoft Entra ID works only with OpenVPN and the Azure VPN Client and adds Conditional Access and MFA; RADIUS forwards authentication to your server and lets you use your own PKI without uploading root certificates to Azure.
- With RADIUS, the route-based VPN gateway forwards authentication to your server (for example NPS), where the gateway is registered as a RADIUS client using the GatewaySubnet and the same shared secret. You can specify two servers for high availability. If the server is on-premises, you need a site-to-site VPN: ExpressRoute can't be used to reach it, and the Basic SKU doesn't support RADIUS.
- Microsoft Entra ID authentication uses OpenVPN and the Azure VPN Client and is configured on the gateway with the Tenant, Audience and Issuer values. With the Microsoft-registered app, nothing needs to be registered in the tenant. Each gateway supports a single Audience; to restrict access by group you use one custom audience app per gateway.
Design, implement, and manage Azure ExpressRoute
- ExpressRoute offers four models: CloudExchange colocation, point-to-point Ethernet, any-to-any (IPVPN) and ExpressRoute Direct. With Layer 2 providers you configure BGP routing yourself; IPVPN providers usually manage Layer 3. Direct gives dedicated, physically isolated ports, with optional MACsec between your routers and Microsoft's.
- Circuits come in three SKUs: Local (one or two regions near the peering location, data transfer included, no Global Reach), Standard (all regions in its geopolitical area) and Premium (global connectivity and Microsoft 365). The billing family is MeteredData or UnlimitedData; you can upgrade from Standard to Premium, but you can't move from UnlimitedData to MeteredData.
- Each circuit has two active connections; keep them on separate on-premises devices and use zone-redundant gateways and BFD. For disaster recovery, use circuits in different peering locations and prefer one with more specific routes, connection weight or AS path prepending while keeping traffic symmetric. Outside the geopolitical region you need Premium; a site-to-site VPN can act as an active-passive backup.
- Global Reach links the private peerings of two circuits to connect on-premises sites over Microsoft's network, using a dedicated /29 subnet and Premium only across geopolitical regions. FastPath sends traffic straight to VMs, bypassing the gateway, and needs SKUs such as ErGw3Az; with Direct it also supports same-region peering and UDRs. Direct provides dedicated ports with MACsec.
- Azure private peering extends your network to virtual networks and their private endpoints over private addresses. Microsoft peering reaches Microsoft 365 and Azure public services by using registered public IPs, usually with NAT, and some services such as Front Door aren't supported. You can enable one or both on the same circuit, each with its own BGP sessions.
Design and implement an Azure Virtual WAN architecture
- A Basic virtual WAN contains only Basic hubs, limited to site-to-site VPN, with no transit router and no scale unit adjustment. Standard adds ExpressRoute, point-to-site User VPN, VNet-to-VNet and hub-to-hub transit, Azure Firewall and NVAs in the hub. You can upgrade from Basic to Standard, but never revert to Basic.
- Virtual WAN implements a hub-and-spoke global transit network: the regional hubs are connected in a full mesh and provide any-to-any transit between branches, remote users and virtual networks, even from another region. Each virtual network connects to only one hub. Branch-to-Branch can be disabled, and traffic is inspected with Azure Firewall or NVAs through routing intent.
- A virtual hub is a Microsoft-managed virtual network with a router that handles transit; it can be created empty and receive its gateways later. Its address space cannot be changed after creation and must not overlap other hubs, on-premises or connected networks; size it with Azure Firewall or NVAs in mind. If routing shows Failed, use Reset Router.
- Scale units set the aggregate throughput of each gateway in the virtual hub and are configured separately for site-to-site, User VPN and ExpressRoute; site-to-site is sized from the total of all branches. Do not confuse them with the routing infrastructure units of the hub router, which you raise for more virtual machines. Changing them can cause brief reconnects.
- Site-to-site, User VPN and ExpressRoute gateways are created with the virtual hub or later by editing it, and they replace virtual network gateways, which must be removed before connecting those networks to the hub. Local, Standard and Premium circuits can connect. Propagate Default Route only forwards a default route that the hub has already learned, for example from a firewall.
3. Design and implement application delivery services
15–20%Design and implement Azure Load Balancer and Azure Traffic Manager
- Azure Load Balancer works at Layer 4 with TCP and UDP. Its components are the frontend IP (public or private), the backend pool, TCP, HTTP or HTTPS health probes, load-balancing rules (inbound only, with optional source IP affinity), HA ports on internal load balancers, inbound NAT rules for port forwarding and outbound rules for SNAT.
- Use Azure Load Balancer for Layer 4 TCP or UDP traffic across VMs or scale sets: public for internet traffic and internal for private traffic, including from on-premises. It also provides SNAT outbound internet access, port forwarding and NVA scale-out with HA ports. Layer 7 needs Application Gateway or Front Door, and Traffic Manager works through DNS only.
- The current SKUs are Standard and Gateway; Basic is retired. Standard supports zone-redundant public or internal frontends, HTTPS probes, HA ports on internal frontends and outbound rules, and has a public-only Global tier that places an anycast IP in front of regional load balancers (not upgraded from a regional one). Gateway inserts NVAs transparently and is chained to a public frontend.
- The frontend type decides the load balancer type: a public IP creates a public load balancer and a private IP creates an internal one, which serves only the virtual network and connected networks such as on-premises over VPN or ExpressRoute, and never accepts internet traffic. Standard load balancers are closed by default and need an NSG that allows traffic. A backend VM can't reach the frontend of its own internal load balancer.
- A regional load balancer spreads traffic across VMs in one virtual network; the Global tier of Load Balancer adds a static anycast public IP in front of public regional load balancers in several regions, preserves the client IP and removes a failed region from rotation. It's created new (a regional one can't be upgraded), supports no outbound rules or internal frontend, and its backend port must match the regional frontend port.
Design and implement Azure Application Gateway
- Application Gateway v2 is a layer 7 load balancer with TLS termination and end-to-end TLS, autoscaling, zone redundancy, a static VIP, URL path and host-based routing, redirection, header and URL rewrite, cookie-based session affinity, connection draining, custom error pages, WebSocket and HTTP/2, and WAF on the WAF_v2 tier. Each requirement maps to one of these capabilities.
- Application Gateway fits when a regional web app needs layer 7 decisions inside the virtual network: path or host-based routing, TLS termination, WAF or cookie affinity, with a public or private frontend for internal apps. It also serves as AKS ingress through AGIC and supports on-premises backends with IP connectivity. UDP belongs to Load Balancer, and global distribution to Front Door or Traffic Manager.
- Application Gateway v2 scales in autoscale or manual mode. Autoscale adds and removes instances with traffic, with a required minimum that keeps baseline capacity while new instances are provisioned and an optional maximum. Manual mode fixes an instance count that doesn't change and can lose traffic when exceeded, so it's sized to the peak plus a buffer. Either way, the dedicated subnet needs one IP per instance.
- An Application Gateway backend pool can hold NICs, scale sets, public or internal IPs, FQDNs and multitenant services such as App Service, inside or outside the virtual network when there's IP connectivity (peering, VPN or ExpressRoute). FQDNs resolve through the virtual network's DNS, and after changing its servers the gateway must be stopped and started. App Service without a custom domain needs a host name override.
- Application Gateway probes every combination of server and backend setting from all its instances and stops sending traffic to unhealthy servers. The default probe uses the protocol and port of the backend setting, host 127.0.0.1, and accepts codes 200-399. A custom probe lets you set host, path, port, interval, threshold and status code or body match criteria; MinServers should be used with caution.
Design and implement Azure Front Door
- Azure Front Door is a CDN and global layer 7 load balancer for HTTP and HTTPS only. It offers four origin routing methods (latency, priority, weighted and session affinity), health probes per origin group, caching, compression, a rules engine, WAF and managed certificates. It supports origins in Azure, on-premises or in other clouds as a Custom host.
- Front Door is the choice for public multi-region HTTP(S) web apps that need edge caching, acceleration, fast failover and WAF. It doesn't fit non-HTTP traffic (TCP or UDP), which goes to Load Balancer or Traffic Manager, nor internal-only apps, where Application Gateway fits; its Private Link only secures the Front Door to origin leg.
- Front Door Standard includes static and dynamic delivery, caching, compression, a rules engine with regex and server variables, custom domains and WAF with custom rules and geo-filtering. Premium adds the managed rule set, bot protection, Private Link to the origin and client mTLS. Standard upgrades to Premium without rebuilding, but there's no direct downgrade; classic tiers don't allow new profiles.
- A Front Door profile contains endpoints with a generated azurefd.net domain, routes that associate domains, protocols and path patterns with an origin group, and origins with priority, weight and host header. Matching picks the most specific route: exact path first, then wildcard; with no match it returns 404. Origin groups define health probes and load-balancing settings.
- Front Door terminates TLS at the PoP with a managed certificate (recommended, autorotated when the CNAME points directly to the endpoint) or with your own certificate stored in Key Vault in the same subscription. For end-to-end TLS you forward with HTTPS, and the origin must present a trusted-CA certificate whose subject name matches its host name.
4. Design and implement private access to Azure services
10–15%Design and implement Azure Private Link service and Azure private endpoints
- A private endpoint is a network interface with a private IP from a subnet and must be in the same region and subscription as the virtual network, even if the target resource is in another region. You need one per subresource, it is reachable from peered networks and from on-premises over VPN or ExpressRoute, and it doesn't block public access by itself. Network policies enable NSGs and UDRs.
- To create a private endpoint you specify the subnet, the resource ID or alias, the subresource (group ID) and the approval method: automatic with permissions on the resource, or manual, which stays Pending. A static IP and the network interface name can only be set at creation. DNS is configured next, for example with a privatelink zone linked to the network and a DNS zone group.
- A private endpoint sends traffic only when its connection is Approved; if the creator lacks permissions, it stays Pending until the resource owner approves it. Network policies are disabled by default on the subnet: enable them to apply NSGs or UDRs, and a UDR must be at least as specific as the virtual network space to override the /32 route.
- A Private Link service publishes an app behind a Standard Load Balancer, tied to a frontend IP configuration and in the same region as the virtual network. Before creating it, disable privateLinkServiceNetworkPolicies on the subnet. Visibility limits who can request a connection, auto-approval removes manual approval, and TCP Proxy v2 reveals the real consumer IP.
- With a private endpoint, public DNS returns a CNAME to the privatelink domain and a private DNS zone linked to the network answers with the private IP, so connection strings do not change. Use the recommended zone names, one zone linked to every network that has clients, and a DNS zone group so records are maintained and deleted with the endpoint.
Design and implement service endpoints
- Use a service endpoint when restricting a PaaS service to specific subnets is enough, with no DNS changes or extra cost: traffic uses the Azure backbone with private source IPs. If you need private access from on-premises, a disabled public endpoint, or per-instance protection against data exfiltration, choose a private endpoint instead.
- Service endpoints are enabled per subnet with the resource name of the service (Microsoft.Storage, Microsoft.Sql, Microsoft.KeyVault and so on), for example with az network vnet subnet update --service-endpoints. Enabling one switches routes on every NIC in the subnet and can drop TCP connections; effective routes then show the VirtualNetworkServiceEndpoint next hop.
- A service endpoint policy, associated with a subnet that already has the Microsoft.Storage endpoint, allows only the listed storage accounts (by account ID, resource group or subscription) and denies every other account in any region. It only supports Azure Storage, does not filter on-premises traffic, and the RA-GRS secondary is allowed automatically.
- To restrict a service to a subnet with a service endpoint, set the default action to Deny and add a virtual network rule; rules based on public IPs stop matching because the source becomes private. On-premises access needs IP rules for the public NAT IP addresses, and the joinViaServiceEndpoint permission is required on the subnet.
5. Design and implement Azure network security services
15–20%Implement and manage network security groups
- An NSG holds rules with priorities from 100 to 4096, evaluated from the lowest number until the first match; two rules cannot share a priority and direction. Default rules such as DenyAllInbound cannot be deleted, only overridden. NSGs are stateful, evaluate private IPs after translation, and rule changes affect only new connections.
- An NSG is associated with subnets and network interfaces, always in the same subscription and location as the virtual network. Inbound, the subnet NSG is evaluated first and then the NIC NSG; outbound, the reverse, and both must allow the traffic. Microsoft recommends a single level; az vm create adds a NIC NSG unless you pass --nsg "".
- An application security group groups network interfaces by role (web, logic, data) so NSG rules need no IP addresses. All its NICs must be in the same virtual network as the first one added, its location cannot be changed, and it cannot be deleted while it contains interfaces. Create it with az network asg create or New-AzApplicationSecurityGroup.
- ASG membership is set on the IP configuration of the network interface, with az network nic ip-config update --application-security-groups or with Set-AzNetworkInterfaceIpConfig and Set-AzNetworkInterface. A NIC can belong to several ASGs in its own virtual network and location, and ASG-based rules affect only member NICs.
- NSG rules are evaluated by priority (lowest number first) and the first match stops processing; default rules cannot be deleted, only overridden. Inbound traffic is evaluated by the subnet NSG and then the NIC NSG, outbound in the reverse order. Augmented rules combine several ports and addresses, and service tags and application security groups avoid maintaining explicit IP addresses.
Design and implement Azure Firewall and Azure Firewall Manager
- Azure Firewall is a stateful firewall with inbound DNAT rules, network rules (IP, port, protocol and, in Standard, FQDNs through DNS proxy) and FQDN-based application rules for HTTP, HTTPS and MSSQL, plus FQDN tags, service tags, outbound SNAT and threat intelligence. It does not apply application rules to inbound traffic: inbound HTTP/S filtering is done with Web Application Firewall.
- Azure Firewall has three SKUs. Basic provides stateful filtering, NAT and application FQDN rules, but threat intelligence only alerts and there is no DNS proxy or network-rule FQDN filtering. Standard adds DNS proxy, FQDNs in network rules, web categories and threat intelligence deny. Premium adds outbound and east-west TLS inspection, IDPS and full URL filtering; inbound TLS inspection belongs to Application Gateway.
- A typical design places Azure Firewall in the hub AzureFirewallSubnet, with 0.0.0.0/0 UDRs to its private IP on spoke subnets (gateway route propagation disabled) and routes on GatewaySubnet that send on-premises traffic to the firewall. Forced tunneling requires the Management NIC in AzureFirewallManagementSubnet, and SNAT behavior is tuned with the private IP address ranges setting.
- To deploy Azure Firewall you create AzureFirewallSubnet, the firewall and its IP configuration with a Standard public IP, and a route table with a 0.0.0.0/0 VirtualAppliance route to the firewall private IP on workload subnets. The az network firewall commands require the azure-firewall extension. The Management NIC can be enabled on an existing firewall by stopping and starting it.
- In an Azure Firewall policy, rules are organized in rule collection groups and rule collections that have their own priorities, but DNAT rules are always processed first, then network rules, then application rules, and rules are terminating. DNAT-matched traffic is allowed without network rule processing, application rules cover HTTP, HTTPS and MSSQL, and unmatched traffic is denied.
Design and implement a Web Application Firewall (WAF) deployment
- Azure Web Application Firewall protects web apps from SQL injection, XSS and other attacks through policies with custom rules (processed first: IP, geography, rate limit, size), managed rule sets (DRS/CRS on Application Gateway, Default Rule Set on Front Door) and Bot Manager, plus exclusions for false positives. Front Door Standard supports only custom rules; managed rules require Premium.
- A WAF design chooses Front Door for global protection at the edge or Application Gateway v2 for regional apps, and associates policies globally, per site (listener) or per URI (path-based rule), with the most specific one winning. Start in detection mode, tune from the logs and move to prevention; when Azure Firewall must inspect traffic to the back ends, Application Gateway sits in front of it.
- Detection mode only monitors and logs matching requests to the WAF logs without blocking; Prevention mode takes the rule action and also logs. With anomaly scoring, Prevention mode acts when the summed severities reach the threshold; on Front Door the action can be Block, Log or Redirect. Start in Detection, tune exclusions, then switch to Prevention.
- On Front Door, managed rule sets (Default Rule Set and Bot Manager) are available only on the Premium tier; Standard supports custom rules only. DRS 2.0 or later uses anomaly scoring, inspects JSON bodies and supports exclusions per rule, rule group or rule set. You can disable rules or change their action, and custom rules are evaluated before managed rules.
- Application Gateway WAF v2 offers the Default Rule Set (DRS), based on OWASP CRS plus Microsoft Threat Intelligence rules, and older CRS versions; only one OWASP-based rule set can be active at a time, but it can be combined with the Bot Manager Rule Set. Rules can be disabled or given another action, the latest DRS runs at Paranoia Level 1 by default, and exclusions remove false positives.