Discussions
Discussions and questions regarding Cato Networks and SASEBest Practices
Find Cato Networks best practices, tips, tricks, and other helpful information!
Recent Content
Is there any way to expose/export DHCP logs from Cato SDP clients?
is there any way (events / API) to see DHCP events for our SDP users? Our security vendors (Rapid7 and Defender for Identity) are doing correlation based on DNS and DHCP events and sometimes see SDP addresses as different machines. I have DNS and PTR records updating but am curious if there is any way to expose the DHCP lease events for SDP users. I see those events for other Cato DHCP but not for SDP users in my tenant. Cato has the concept of "User Awareness" that is correlating IP addresses to User IDs. When we were using Windows DHCP servers we fed the logs to our security vendors for a similar type of correlation between IP addresses and User IDs. As we are moving away from Windows Servers in our offices, we are losing this visibility. We are beginning to allow Cato to provide the DHCP on our LAN segments, as well as for our remote SDP client users. As this happens, we are seeing DHCP events on the LAN segments which can be tied to machine names and matched against login events via active directory or Entra ID to correlate IP addresses to users. However, for our remote SDP client users I cannot seem to find DHCP events. This leads to issues. Microsoft Defender for Endpoint sees a user getting different IP remote SDP client addresses in the 10.41.x.x as "Pass the Hash" attacks. However, when I investigate, it is the same workstation being getting different IP addresses through normal, remote operation. If the user does not reboot/login every day this raises security alerts. Am I missing the point, or not configuring something correctly? Is there a way via API or syslog forwarding to monitor DHCP logs from Cato for both LAN segments and SDP client segments? The ultimate solution would be log forwarding type of solution where I could forward all Cato DHCP lease events to Microsoft Defender for Endpoints/Identity and my security vendor (Rapid7) but I am just wondering how others are handling this. I figured I would ask around before I put something in the Idea hub for a non-issue.
Migration SCIM to SCIM Provisioning
Hello Team, We currently use Okta for SSO and SCIM provisioning with Cato. We want to keep SSO authentication on Okta, but move provisioning from Okta to Saviynt. Our Understanding: - We understand SCIM endpoints are scoped per directory (sourceId) as /scim/v2/{accountId}/{sourceId}. - We also noted documentation stating multiple IdPs are supported, but not recommended as a migration method. Could you please advise the following? Is there a recommended procedure to migrate existing users from an Okta SCIM directory to a new Saviynt SCIM directory? If adding a new SCIM directory is not recommended for migration, please point us to the recommended migration steps or best practices. We want users and groups provisioned via Saviynt to authenticate via Okta SSO. Is mapping Saviynt directory to the Okta SSO provider a possible configuration? Please let us know if you have any recommend plan. Thank you,Degraded Sockets in High Availability
I have multiple customers that have a LTE sim card just for the main socket. This will have the sockets identify asymmetric WAN connections causing the DEGRADED alert. What can I do to disable the DEGRADED alarm from the site? could it be possible to disable the interfaces so the asymmetric connections don't show as alarmed?
LAN Firewall rules - missing "IP range" in src/dst
Anyone else missing an ability to use Custom IP Range as a source or destination in LAN Firewall rule? We use CATO LAN Firewall to control traffic between two separate network zones terminated on two different internal firewalls. Since this is a local traffic in the site, we don't want to route it to Cato Cloud so it's not dependent on WAN links. That's why we use CATO LAN Firewall (formerly Local Routing). But the only options to set Source or Destination are: Global range, Host, Interface subnet, Network Interface and Any. Would be very useful if we can use Custom IP ranges and Host Groups there.
Pre-Login and Online Services
We currently have an on-premises Active Directory and have Pre-Login enabled with connect at boot enabled. We defined internal destinations (domain domain controllers) as allowed destinations, so the devices can reach the domain controllers before the user has logged in. This worked fine so far. However, now we want to migrate to Entra ID and Intune only, which means that the machines now need to reach Entra and Intune before or directly after the login. Since the pre login mode doesn't allow them to reach all URLs of Entra ID and Intune, we get problems during log in and for the Intune enrollement (which happens after the login of a new user but before the user has authenticated with the CATO client). We also have the same problem with NinjaOne which we use to manage endpoints: We would like to be able to reach endpoints before a user has logged in. In the allowed destinations for the Pre login mode, I can only provide internal targets and IPs, but can't put any Internet hostnames so the devices can reach Entra ID and Intune before the user has authenticated. So what is the solution here? We want to use Pre login to have the security it provides and prevents the devices from having open Internet access before the user has authenticated with CATO, but really need to resolve these issues that are caused by it when it comes to connect to our management services before the user has authenticated. Thank you in advance.
Container based FQDN addition- Anomaly behaviour
Dear All, I get malicious domain and IPs to block from my regulator. Instead of manually adding IPs and domain, I created Container and than get it sync with Raw Github for both IPs and FQDN. The FQDN and IPs are syncing correctly in CMA. To test, I added "linkedin.com" to malicious domain in container and created Internet Firewall rule to block for this container. When I browse "linkedin.com", I get block page but as soon as I type "www.linkedin.com", it opens and not get blocked. Since there is only option of either IP or FQDN to be added in Cato Container, now due to this limitation, my purpose is not solved. Ideally it should block all subdomains as well but it only blocks if FQDN is browsed. So with this, malicious attacker URLs will not be blocked. Cato should provide option to add "Domain" as well along with IP and FQDN as this is not solving the problem of blocking malicious domain basis threat feeds received from regulator. Cato Product should look into this.Recording: Ask Me Anything with Professional Services - March 2026
Professional Services AMA – March 2026 Thank you to everyone who joined our March AMA session! Below is a clean, easy‑to‑scan recap of every question asked, along with brief summaries of the answers shared during the call. If you’d like the full context, you can view the recording below. IPsec & Tunnel Behavior Are IPsec tunnels active‑active or active‑passive? Cato uses active‑passive by default. Active‑active is available with EA enablement. How do we identify the primary vs. secondary tunnel? Your IPSec configuration labels them directly: Primary=active Secondary=standby Can secondary tunnels load‑balance with primary? No. Secondary tunnels do not participate in load‑sharing. Can we configure active‑active instead of active‑passive? Yes, but only via Early Access. This enables multiple tunnels on primary/secondary POPs. Logging, Packet Capture & Troubleshooting Can we see packet counts like a firewall (sent/received)? You can see traffic hitting WAN firewall policies, but tunnel‑level packet breakdown is not available. Can we capture tunnel traffic? Not in CMA. Support can perform captures on the backend. Bandwidth, Utilization & Alerts Is there an automated way to detect when a site hits max bandwidth? Yes, via the API or MCP Server integrations. Are bandwidth‑based alerts planned in CMA? Not today. Current alerts may trigger on QoS discards. Recommended to submit an Idea Hub request. Can we mute all alerts during a maintenance window from one place? Not currently. Alerts must be disabled per area (BGP, XOps, link health). Suggested as an idea via the Idea Hub. Browser Extension, DNS & Clientless Access Hostname access fails with the browser extension. Why does IP work? Hostname resolution should work in normal circumstances. Re‑testing and a support ticket is recommended if it persists. Direct Connect, IPv6 & Packet Capture When will packet capture be supported on Direct Connect? No timeline yet; currently only possible on the backend through Support. AI Security Monitoring Does AI Security capture user prompts (e.g., Copilot prompts)? Robin gave a walkthrough of our AI Security offering during this event. He starts discussing capturing user prompts around the 18th minute and continues discussing how to secure and monitor AI for several minutes. AWS Architecture & Inspection Can inbound AWS traffic be inspected by Cato before reaching EC2 (like GWLB + Palo Alto)? Not with AWS public IPs. Cato can only inspect inbound traffic terminated on Cato public IPs via Remote Port Forwarding. Why must inbound inspection use Cato’s public IP? Cato is a SaaS platform and cannot locally inspect traffic inside your AWS VPC. Automation, Importing & Configuration Management Is there a bulk import feature for IP ranges/VLANs? Not natively. Consider: Cato API CatoCLI Terraform provider Best Practices & Identity What best practice do customers commonly miss? Fully adopting identity‑based policies (ZTNA) instead of legacy IP‑based access controls. Do AD‑synced users need a ZTNA/SDP license for identity policies? It depends: Windows + SCIM + Azure AD Join/Hybrid = no license needed macOS = license currently required On‑prem AD join = SCIM not supported (use LDAP) Why does user awareness fail for some SCIM‑synced users? SCIM does not support on‑prem AD joined devices. These must use LDAP provisioning. Always‑On VPN Issues Why does always‑on VPN block all traffic until reinstall? Common causes include: Internet Recovery option not enabled Device posture checks failing If issues persist, Support should investigate. Event Logs Can we filter traffic to wildcard domains? Yes, use the “contains” filter for domain‑based event searching. Remote Browser Access What’s the high‑level architecture for Remote Browser Access? User connects to the Cato Portal Portal creates a policy Cato initiates a connection to your internal resource Without Source NAT, the internal server sees a Cato public IP. Source NAT forces it to appear from a private IP instead. QoS for Remote Port Forwarding Can we set QoS rules for Remote Port Forwarding? Not today, traffic uses the default QoS queue. Idea Hub submissions encouraged. Local Bypass Enhancements Will more applications be added to local bypass? Yes. The list is expanding, and domain/FQDN bypass is available in EA via your account team. Questions Requiring Follow‑Up These topics require SME confirmation and will be answered on the community once available: 1. Does AI Security capture user prompts (Copilot, etc.)? Pending SME validation. 2. Is IPv6 DNS fully supported, and how does Cato plan to address IPv6‑only ISP environments? Pending SME validation. Have more questions? Drop them in the community anytime or join our next AMA.
