Predefined EDLs are not downloaded on newly onboarded Explicit Proxy regions. This condition occurs only if you configure and onboard new Explicit Proxy regions using XML APIs or CLI.

Workaround: Select PanoramaCloud ServicesConfigurationMobile Users—Explicit Proxy, click the gear to edit the Settings, and click OK to apply the predefined EDLs to the Explicit Proxy configuration.

After upgrading to the 3.2.1 Cloud Services plugin, the options to configure Mobile Users - GlobalProtect IP Address Pool at a per-location group do not display; however the choices to configure an IP Address pool at a Worldwide or Regional level display.

Workaround: Log out and then log back in after upgrading the plugin for the per-location group IP pool options to display.

When using a Panorama running 11.0 to manage Prisma Access, logins to Panorama Managed Prisma Access take a long period of time (two minutes or more).

When using a Panorama with a version of 11.0 to manage Prisma Access, you cannot use the Edit Selections option during Commit and Push operations or all windows will close.

Workaround: Do not use change the Edit Selections options, or uninstall the Cloud Services plugin.

When performing commits or upgrades, a Prisma Access deployment requires internet connectivity; without internet connectivity, certificate validation will fail and commits are not possible.

In Prisma Access Explicit Proxy deployments, different device connections may show identical entries in the Authentication logs if the source IP address and browser version are the same for both device connections.

When upgrading your Panorama from an earlier 10.1 version to 10.1.7 and you use HIP profiles, local commits fail with 'hip-profiles unexpected here' and 'rules is invalid' errors. This condition is the effect of HIP-profile objects in security policies and authentication policies being replaced with source-hip and destination-hip objects.

Workaround: Start a CLI session with the Panorama that manages Prisma Access and enter the following commands:

When onboarding an Inbound Access Remote Network, commits fail with a Commit-all Error: Saas-Agent Exception: failed to get SPN error. This is an intermittent issue.

Workaround: Delete the Inbound Access Remote Network from Panorama and perform a Commit and Push operation; then, re-add the Inbound Access Remote Network and perform another commit and push operation.

If you use a Panorama version of 10.2.3 or later to manage Prisma Access, certificates can be lost after a plugin upgrade, which causes an OTP request to be issued.

If you use a Panorama running PAN-OS version 10.1.7 to manage Prisma Access, attempting to access the online help in the Cloud Services plugin area (PanoramaCloud Services Configuration or PanoramaCloud Services Status), the online help in the 3.2 Cloud Services plugin does not display.

Workaround: Upgrade your Panorama to 10.1.7-h1 to view the online help.

If you have remote network locations that were remapped to new compute locations as part of the 3.2 infrastructure upgrade, you receive a null message when attempting to allocate the bandwidth for the remapped compute locations.

Workaround: Open a CLI session with the Panorama that manages Prisma Access and enter the debug plugins cloud_services prisma-access refresh-infra-files command to refresh the location-to-compute location mapping.

Advanced Threat Protection (ATP) is not supported when using TLS 1.3.

Authentication override values in portal and gateway configurations are not accepted when the following conditions apply:

If your deployment meets these conditions, authentication cookie overrides are not applied (the Generate cookie for authentication override and Accept cookie for authentication override values in portal and gateway configurations are not accepted).

While browsing through various tabs under in the Panorama UI under PanoramaCloud Services, a blank pop-up window might display with a title of Error.

Workaround: This issue has not been found to create any functional impact. Closing the window and refreshing the UI should solve the display issue.

While browsing through various tabs under in the Panorama UI under PanoramaCloud Services, a blank pop-up window might display with a title of Operation Failed.

Workaround: This issue has not been found to create any functional impact. Closing the window and refreshing the UI should solve the display issue.

If you have QoS enabled in a Remote Network compute location, and you reduce the bandwidth for a compute location enough that Prisma Access removes an IPSec termination node from that compute location, QoS is disabled for that compute location.

Workaround: After the IPSec termination node has been deprovisioned, re-enable QoS for that compute location.

When using a Panorama running 10.2 to manage Panorama Managed Prisma Access, Inactivity-Logout values can only be configured using minutes.

If you are managing an on-premise or VM firewall running 10.0 with a Panorama running 10.1 or 10.2, an Inactivity-Logout configured on Panorama is configured as disconnect-on-idle on the managed firewall.

Local commits to Panorama are not able to be performed during a Prisma Access maintenance window.

If a GlobalProtect mobile user has a dollar sign ($) in the user name, they cannot log out of Prisma Access.

Workaround: Do not use dollar signs in GlobalProtect user names.

Onboarding and autoscaling of Mobile User locations is successful even though the Mobile User IP address pool is insufficient to onboard the locations or allow autoscaling events.

If you are using a Panorama with a version of 10.2 or later to manage Prisma Access and you specify Prisma Access to append the ending token to URLs in URL filtering configuration under PanoramaCloud ServicesConfigurationService SetupSettingsAdvanced, this setting might differ from the Append Ending Token setting in the DeviceSetupURL Filtering area.

Workaround: Make sure that the two values are the same in Panorama.

When using a 10.2.2 Panorama to manage a Panorama Managed Prisma Access 3.1.2 deployment, when attempting to download Preview Rules in the Mobile_User_Device_Group (PoliciesPreview RulesPDF/CSV), a 500 Internal Server Error is received.

When using the South Africa West, France North, Ireland, Bahrain, or South Korea Explicit Proxy locations, mobile users have difficulty connecting to some websites.

Workaround: Deactivate these locations and use any of the Explicit Proxy supported locations.

After an upgrade from the 2.2 Preferred Cloud Services plugin to 3.x, the Troubleshooting Commands (PanoramaCloud ServicesConfigurationService SetupTroubleshooting Commands) failed to display the Logging status, Routing information, EDL info, EDL status, EDL refresh, and Search EDL fields.

When onboarding a remote network and selecting options such as Summarize Mobile User Routes before advertising, Enable BGP, Don't Advertise Prisma Access Routes, or Advertise Default Route, an Object already exists error is displayed.

Workaround: Cancel the current onboarding attempt and retry the operation. This error is transient and subsequent retries should not experience this issue.

Traffic that matches the intrazone-default and interzone-default rules do not display in the logs.

Workaround: This traffic does not display in the logs by default. To have them display, create rules for the intrazone and interzone traffic.

If you have enabled cloud provider redundancy for service connections, the Redundancy Assessment area in the Network Details tab (PanoramaCloud ServicesStatusNetwork DetailsService Connection) shows (link to published locations).

When, in an Explicit Proxy deployment that does not have Remote Networks onboarded, you select Forward Remote Network traffic to Explicit Proxy in the Advanced tab, the first three octets of the IP addresses display as None (for example, None.254 instead of 172.25.255.254).

Workaround: Onboard a remote network and Commit and Push your changes, making sure that both Explicit Proxy and Remote Networks are selected in the Push Scope.

If you have QoS profiles with a Class Bandwidth Type of Mbps, validation fails and you receive the following error:

For QoS profile profile-name , summation of its class egress-guaranteed is Mbps-value , which is larger than its egress-max Mbps-value.

This error displays if all of the following conditions are true:

If you have an Egress Guaranteed value of 0 in the Profile area, the summation of the Egress Guaranteed values in the Classes field cannot exceed the Egress Max value in the Profile area.

This restriction applies to all QoS profiles in the template stack, even if they are not being used.

Workaround: Delete the profile, or modify any QoS Profiles so that the summation of the Egress Guaranteed values in the Classes field does not exceed the Egress Max value in the Profile area.

In Prisma Access 3.2.1, mobile user IP Address pools and you specify IP addresses at a location group level, the IP addresses in the pool do not get released after the mobile user disconnects from GlobalProtect.

Workaround: Remove the IP address pool at the location group level and use Worldwide or Regional pool addresses only, or reach out to your Palo Alto Networks account representative or partner, who will contact the SRE team to release the IP pools.

If you onboard service connections using the Cloud Services plugin 3.0 in multitenant mode, you cannot view the service connections in the drop-down list if you perform the following actions:

Workaround: Do not load a configuration from a previous plugin version after upgrading to a newer plugin version. The configuration load also causes the previous plugin version to be loaded, which is an unsupported configuration.

When changing the Local IP Address in the BGP tab for a Remote Network connection that uses BGP, the following issues can be seen:

Workaround: Refresh the Panorama UI. If a refresh does not fix the issue, change the Local IP Address to a placeholder value, click OK, and then re-enter the correct Local IP Address.

When a new Explicit Proxy instance is created, the threat logs may not send device group information. This behavior can occur in a new deployment or can change in an existing deployment after a maintenance activity or infrastructure upgrade.

Workaround: Select All instead of a specific Device Group when viewing logs.

After successfully completing a partial commit, the Commit Status messages includes the message Changes to all template configuration.

Workaround: Ignore the message regarding all templates being changed. The partial commit was performed only for the template or Commit Scope you specified.

After migrating from a remote network deployment that allocates bandwidth by location to one that allocates bandwidth by compute location, QoS statistics are not displayed for inbound access sites.

If you use the remote network aggregate bandwidth model and you enable QoS for a remote network (PanoramaCloud ServicesConfigurationRemote NetworksSettingsQoS) that has ECMP enabled, you must select Customize Per Site and click OK or you will receive an error on commit.

When you log out mobile users from the PanoramaCloud ServicesStatusStatusMobile Users—GlobalProtect area using the Logout function, or if you log out a user using CLI, the user is successfully logged out, but the Current Users area might still show the user as being logged in for up to five minutes after the logout activity occurred. This behavior is the result of Prisma Access refreshing the status of logged-in users every five minutes. If you have configured a Connect Method of User-logon (Always On) or Pre-logon (Always On), and if the user reconnects during the five minute refresh interval, the user might not be reflected as being logged out in the Current Users area.

Workaround: View the login and logout events from the GlobalProtect logs.

In a multi-tenant environment, you cannot enable the EDL Custom Category End Token Support feature until all your tenants have had their infrastructure and dataplane upgraded to meet the requirements for the 3.0 Cloud Services plugin.

Workaround: Wait until all your tenants have had their infrastructure and dataplane upgraded before enabling the EDL Custom Category End Token Support feature.

When viewing the Push State Details after a commit to a device group, you see a message similar to Interface tunnel.2 has no zone configuration’.

Workaround: This is a spurious message related to a backup tunnel configuration and can be ignored.

When using Traffic Steering, when a user matches a URL in an EDL, pre-defined URL category, or Custom URL Category, the first two sessions are not directed to the target for internet-bound traffic.

You cannot make any configuration changes in the Advanced tab under Explicit Proxy Settings (PanoramaCloud ServicesConfigurationExplicit ProxySettingsAdvanced).

Workaround: There is no workaround. This functionality will be supported in a future Prisma Access release.

When using the Egress IP Allow List feature in Prisma Access, you might experience the following issues when using the UI:

If you install an Innovation release, configure a feature that is only supported on an Innovation release, and then migrate from an Innovation to a Preferred release, you receive a commit validation error after making configuration changes in the Cloud Services plugin.

Workaround: Delete the unsupported feature by creating a CLI session with the Panorama that manages Prisma Access in configuration mode and entering the delete plugins cloud_services <feature-name> command, where <feature-name> is the name of the feature that is unsupported in the Preferred release.

When using the Enterprise DLP plugin with Prisma Access, an uploaded file that matched a Block action on a data filtering profile was not blocked from being uploaded, along with an error DLP Skipped: missing boundary m in the Data Filtering logs.

When configuring QoS for remote networks (PanoramaCloud ServicesConfigurationRemote NetworksSettingsQoS), you can select None as a QoS Profile.

Workaround: Select a valid QoS profile to enable QoS. None is an invalid selection.

When viewing logs for an Explicit Proxy deployment, duplicate log entries might be seen. This behavior does not affect Prisma Access functionality.

If you are configuring a Mobile User - GlobalProtect deployment, if you do not enable the allow listing feature when configuring or onboarding the mobile user deployment, the plugin logs might display spurious messages that are similar to the following messages:

2022-01-13 13:14:27.217 -0800 INFO: [access-domain-xpaths] Sending result back <result><status>pass</status><msg>cloud_services</msg><msg>cloud_services/access-domain</msg></result>2022-01-13 13:14:27.290 -0800 ERROR: [get_ip_allowlist_addresses] yes-allow-list node not found! Please config yes-allow-list under ip-allow-list node.

Workaround: Ignore the plugin messages; these messages do not affect normal Prisma Access operation.

If, when in an Explicit Proxy deployment that is forwarding remote network traffic to Explicit Proxy, if you deselect the Forward Remote Network Traffic to Explicit Proxy check box in the Advanced tab, the IP addresses that were allocated by Explicit Proxy still display in the Advanced tab.

Workaround: Refresh the Panorama UI to clear the IP addresses in the UI.

When Prisma Access creates a new compute location and remaps an existing remote network location to that new location, if you do not delete and re-add the existing compute location to take advantage of the latest compute location-to-location mapping, you cannot view bandwidth statistics for the remapped location.

Workaround: Delete and re-add the remote network location that is associated with the new compute location. The Service IP Address will change, so you will have to change the IP address for the IPSec tunnel on your CPE to the new Service IP Address, and you will need to commit and push your changes twice (once after you delete the location, and once after you re-add it).

When configuring the IP addresses to use to forward remote network traffic to Explicit Proxy in the Explicit Proxy Advanced settings (PanoramaCloud ServicesConfiguration Mobile Users—Explicit ProxySettingsAdvanced), Remote Networks does not display in the Push Scope for a Commit and Push operation.

Workaround: Select Remote Networks as well as Explicit Proxy in the Push Scope before performing a Commit and Push operation. Forwarding traffic from remote networks to Explicit Proxy requires that you commit and push changes to both Explicit Proxy and remote networks.

Strata Logging Service failed to reconnect after a disconnect if a management IP address used for logging had an IP address assignment type of DHCP.

When you run the API to retrieve Prisma Access IP addresses with a serviceType of all, the API times out if your deployment has a large number of Remote Networks.

Workaround: If you have a large number of remote networks, specify a serviceType of remote_network instead of all when running the API.

If you have created a remote network deployment that allocates bandwidth by compute location and then delete the remote network license, any commit for changes to features that are still licensed fail with an Failed plugin validation error.

Workaround: Delete the unused remote network configuration by opening a CLI session with admin-level privileges, entering configure to enter configuration mode, and then entering delete plugins cloud_services remote-networks. Then, retry the commit operation.

If the dataplane is not compatible with the plugin you are running, a generic message indicating that the Panorama is undergoing maintenance displays in the Panorama Alert and Plugin Alert fields in PanoramaCloud ServicesConfigurationService Setup.

When completing a mobile user setup in a FedRAMP Moderate deployment and configuring the mobile user IP address pool, you receive an Operation Failed message with text that indicates that Prisma Access could not auto-generate an authentication cookie certificate. In addition, when committing and pushing your changes, you receive a validation error related to a cookie decryption certificate.

Workaround: Create a signed certificate and apply it to the Mobile Users—GlobalProtect configuration by completing the following steps:

If you are using a Panorama of a version or 10.0 or lower, and you configure an invalid destination port value anywhere in Panorama (for example, in ObjectsServices), a commit-all operation fails with a vague error related to a module or device having a Non digit value.

Workaround: Fix the invalid port configuration, then retry the commit-all operation. Panoramas running 10.1 or later disallow you from configuring an invalid destination port value.

When you Enable IPv6, a window displays asking you to enable Telemetry Data Collection.

Workaround: Click Remind Me Later to dismiss the window.

If you have applied QoS to your remote network deployment but have not yet committed and pushed your changes, the QoS statistics screens display blank information.

Workaround: Commit and Push your QoS changes for the QoS statistics to display.

If, when using Explicit Proxy, when the following conditions exist, mobile users might experience issues with CORS requests and non-decrypted traffic:

Workaround: Clear your browser's cache to re-authenticate with the ACS.

BGP addresses ending with .0 or .255 are not allowed to be entered in the UI as peer BGP addresses for service connections or remote networks, regardless of the subnet being used.

Workaround: Use CLI commands to enter the .0 or .255 address by logging in to the Panorama that manages Prisma Access and entering one of the following commands:

set plugins cloud_services service-connection onboarding sc-name protocol bgp peer-ip-address ip-address

set plugins cloud_services remote-networks onboarding rn-name protocol bgp peer-ip-address ip-address

Where sc-name or rn-name is the name of the service connection or remote network connection.

If you have IPv6 enabled in your Prisma Access deployment, the Private IPv4 address of mobile users (PanoramaCloud ServicesStatusMonitorMobile Users—GlobalProtectLocationsUsers) is displayed, but the IPv6 Private IPv6 address of mobile users is not.

IP precedence-based classification is not working for Prisma Access, when using either IPv4 or IPv6 IP addresses.

When you enable IPv6 for a single tenant in a multi-tenant deployment, the UI page refreshes and displays the Cloud ServicesConfiguration page, where you select the drop-down for all tenants.

When configuring mobile users DNS settings in the Network Services tab, you should not enter Custom DNS Server IP addresses (either IPv4 or IPv6) without also specifying a Domain List.

Workaround: Specify a Domain List.

If you add an IPv6 address pool to your Mobile Users—GlobalProtect deployment, select the regions to Enable IPv6 in the IPv6 Availability tab, and Commit and Push your changes, the pools appear in the IPv6 Availability tab. If you then disable all regions, effectively disabling IPv6, and then Commit and Push your changes, the IPv6 address pools still display in the IPv6 Address Pool tab.

Workaround: There is no workaround. If you later enable IPv6 for one or more regions, you can use the existing IPv6 address pool. You can also specify a different IPv6 address in the IP Pools and, after you commit and push your changes, the new IPv6 Address pool overwrites the existing addresses and displays in the IPv6 Availability tab.

When viewing or changing QoS settings for Remote Networks in Panorama Cloud ServicesConfigurationRemote NetworksSettingsQoS, a newly-added compute location or location does not display.

In addition, a newly-onboarded location does not display in the Site Allocation (Customize Per Site) page.

Workaround: Refresh the Panorama that manages Prisma Access.

In a multi-tenant deployment, you receive a Configuration committed successfully message along with a Not all Commit-All jobs got triggered message.

Workaround: Either upgrade your Panorama to a minimum version of 10.1.4, or select CommitCommit and Push, Edit Selections, and in the Prisma Access tab, make sure that the Push Scope includes the changes you made for the Prisma Access configuration. Depending on the changes you made, select one or more of the Remote Networks, Mobile Users, Service Setup, and Explicit Proxy choices.

If you are sinkholing IPv6 traffic, the policy rule hit counts for traffic that matches the IPv6 sinkhole policy do not increment when entering the CLI command show rule-hit-count vsys vsys-name vsys1 rule-base security rules all.

In a multi-tenant deployment, admin users that have more than one access domain cannot configure new remote networks or service connections, and can only view what is already deployed.

Workaround: Create the access domain first, then select the access domain you created when you convert the single tenant to a multi-tenant setup.

When you select Integrate with Prisma SD-WAN, the integration fails.

When downloading a large file (including but not limited to programs, browser extensions, or apps) using Explicit Proxy, if the download takes longer than the cookie lifetime, the download fails when the cookie expires.

If, after signing in to Explicit Proxy, you open a link that contains a file to download, the file downloads successfully but the Explicit Proxy sign-in page continues to display.

Workaround: Since the link contained a downloaded file, there is no page to display and the current page does not refresh. Select another webpage to navigate away from the sign-in page.

If you are using a Panorama with a version of PAN-OS 10.1 to manage Prisma Access, and you migrate a Remote Network deployment from allocating bandwidth by location to allocating bandwidth by compute location, the migration banner displays the location names in an incorrect (large) font.

Workaround: No workaround is required. There is no change to the migration functionality; the only issue is with the font displayed during the migration.

When using Troubleshooting Commands (PanoramaCloud ServicesConfigurationService SetupService OperationsTroubleshooting Commands) with Panoramas that are in High Availability mode, the commands cannot be run from the passive Panorama.

If you delete an explicit proxy configuration and then reconfigure it within 10 minutes of its deletion, Prisma Access cannot properly process the new configuration and explicit proxy functionality could be affected.

Workaround: Wait at least 10 minutes after deleting an explicit proxy configuration before reconfiguring it.

When using Panorama 10.x to manage Prisma Access, if you configure an Authentication Enforcement Profile under ObjectsAuthentication and specify an Authentication Profile that resides in a Shared location, you receive an error when committing the changes.

Workaround: If you use a Panorama 10.x to manage Prisma Access, do not use a shared Authentication Profile for any Authentication Enforcement Profile; instead, use an Authentication Profile that is under one of the Prisma Access Templates.

When using explicit proxy, there could be a delay when displaying user details under Current User Count due to a log ingestion issue between explicit proxy and Strata Logging Service.

When performing a local commit or Commit and Push operation, you receive the error Internal Server Error: Failed to aggregate bandwidth configuration.

Workaround: Check the DNS configuration of the Panorama appliance that manages Prisma Access, and check that Panorama is able to contact your network's DNS servers, then retry the operation.

If, during Explicit Proxy onboarding, you onboard a large number of locations, the Explicit Proxy status might display its status incorrectly (for example, a status of ERROR might display when the onboarding was successful).

If you change the Explicit Proxy URL in Prisma Access but do not change the PAC file to reflect the change, the change won't be applied.

Workaround: Upload a new PAC file with the same changes as you made in the Explicit Proxy URL.

If you change the proxy FQDN, the changes are not immediately reflected after the job status completes.

Workaround: Workaround: Wait 10 to 15 minutes for the changes to be reflected after the Job status shows as Completed on Panorama.

There is a delay observed to populate the Rule Usage column on the Policies page.

Workaround: Refresh the page by clicking on the refresh button on the right side.

In addition, the Preview Rules tab does not display the Rule Hit counters.

Workaround: Click the Used link on Rule Usage column to display the Rule Hit count for the rule.

The maximum length of a URL that can be used with explicit proxy is 1280 characters.

WildFire logs show explicit proxy logs as having a source zone of Proxy. If you use a name of Proxy for Clean Pipe instances or remote networks, you will not be able to differentiate between explicit proxy logs and logs with the clean pipe or remote network name of Proxy.

Workaround: If you use explicit proxy, do not specify a name of Proxy for any Clean Pipe instances or remote networks.

The PanoramaCloud ServicesStatusMonitorMobile UsersExplicit Proxy page incorrectly shows the current number of users as 0.

When using Explicit Proxy, initial DNS Queries (first leg) and Initial HTTP connect messages (first logs) are not seen in the traffic logs in Panorama.

When you enter the show pbf extended-address all command to retrieve the traffic steering cache, an FQDN displays with an asterisk, such as *.example.com.

Workaround: No workaround is required. The displayed FQDN is correlated to the FQDN server that presented the certificate.

When configuring a Mobile Users - GlobalProtect deployment using SAML authentication, you receive a pangp.gpcloudservice.com is missing certificate error when you commit your configuration changes.

Workaround: Add the missing certificate in your SAML IdP configuration by selecting DeviceMobile_User_TemplateAuthentication Profile in Panorama and adding the certificate.

A webpage may contain links of resources from the domains other than the domain from where the webpage is served. Most modern browsers do not send any cookie along with the requests to get the resources from those third-party domains for security reasons. Since there is no cookie present to identify the user for those third-party domains, the user name cannot be logged in the traffic logs for those domains.

In addition, there will be some connections that Prisma Access redirects for authenticating a user. Logs for such connections will not have any username.

When using traffic steering, if you specify External Dynamic List that has an IP address and port, traffic is not forwarded to the target.

Workaround: Remove the port number from the IP address.

When using explicit proxy, if you update the cookie lifetime to a shorter lifetime than the previously configured value, the new lifetime value does not apply to users who are already logged in until the original longer life time expires. New users logging into the service receive the new shorter cookie life time.

Explicit proxy configuration changes are not applied to the configuration after a commit.

Workaround: If you are not seeing the changes after retrying the commit operation, contact Palo Alto Networks support.

When administrators log out a mobile user who is logged in using SAML from the Prisma Access status page (PanoramaCloud ServicesStatusStatusCurrent Users), a Single Logout (SLO) request is not generated. As a result, the user is logged out of the gateway but is not logged out of the IdP, and if the client SAML cookie is still valid, the user can reconnect without having to input credentials.

Extra IPSec termination nodes are allocated to a compute location if you allocate bandwidth multiple times in a very short time interval.

When you allocate Bandwidth to a compute location from the Onboarding section, that allocation is not reflected immediately in the Bandwidth Allocation tab until you manually refresh the page.

Workaround: Manually refresh the Panorama that manages Prisma Access.

When you upgrade the Cloud Services plugin and then perform a commit operation, not all Prisma Access components are selected in the Push Scope.

Workaround: Select CommitCommit and Push, Edit Selections in the Push Scope, and make sure that all Prisma Access components (Service Setup, Remote Networks, Mobile User, and Clean Pipe, depending on your license) are selected before committing and pushing your changes.

When you change the name of a target service connection group for traffic steering, the updated target name does not display in the Traffic Steering Rules area.

Workaround: Refresh the Panorama browser.

If you use IKEv2 with certificate-based authentication, only SHA1 is supported in IKE crypto profiles (Phase 1).

Workaround: Use an IKEv2 (Phase 1) cryptographic profile of SHA1 on your customer premises equipment and in Prisma Access.

If a service connection loses both its active and backup connectivity, mobile users lose connectivity to users and resources connected to Remote Networks and Service Connections.

If you have two Panorama appliances configured in high-availability mode, the passive Panorama will display an out of sync message during a commit and push operation.

Workaround: Open a command-line interface (CLI) session on both the passive and active Panorama and enter the following commands:

username@hostname> debugmd5sum_cache clear

username@hostname> configure

username@hostname# commit force

Prisma Access bypasses Traffic Steering for rules with a service type of HTTP or HTTPS if you use an application override policy for TCP ports 80 and 443.

In addition, traffic steering does not work for URLs from URL categories referenced in the traffic forwarding rule if you have configured an application override policy for TCP ports 80 or 443.

Mobile user route summarization is not supported in hot potato routing mode.

When using hot potato routing, Mobile User route summarization may add extra latency for traffic between mobile users and headquarters or branch traffic.

After you create a traffic steering rule with an IP address, IP address group, EDL, or custom URL category as a Shared object, make changes to any of those objects, and then commit and push your changes, only the Shared object displays in the Push Scope. Prisma Access device groups doesn't get displayed in the push scope.

Workaround: Select CommitCommit and Push, Edit Selections in the Push Scope, and make sure that you select all device groups (Service Setup, Remote Networks, Mobile User, and Clean Pipe, depending on your license) before committing and pushing your changes.

When adding or deleting URLs to a custom URL category, Prisma Access does not purge its cache, and the change does not immediately take effect.

Workaround: Perform one of the following actions:

To make sure that Prisma Access can distinguish between users if the same username is shared between users who authenticate locally and users who authenticate using LDAP, you should authenticate LDAP users in the format of domain/username and authenticate local users in the format of username (without the domain name).

Do not create any custom URL categories that start with GPCS-, gpcs-. or custom_url_category_pbf.

If Panorama access is disabled in an Admin Role Profile, you can still see the contents of the plugin, but the fields are read-only.

When you upgrade the Cloud Services plugin to 1.7, Prisma Access prepends an asterisk to URLs in custom URL categories, if you use this category in a traffic steering forwarding rule. If you use the same URL category policies for both traffic steering and other security policy rules, these changes apply to both the traffic steering rules and other security policy rules.

If you have custom URL categories that are not used in traffic steering forwarding rules, Prisma Access does not change the URLs in those categories.

Prisma Access prepends an asterisk to URLs in custom URL categories, which doubles the number of URLs entered in a custom URL category. Prisma Access supports a maximum of 300,000 URLs in URL category entries; if you use custom URLs for traffic steering and are close to this limit, the doubling of URLs might cause your deployment to exceed the limit of URLs.

If you used policy-based forwarding rules to forward internet-bound traffic to service connections in Prisma Access 1.6, Prisma Access makes the following additions to URLs in custom URL categories after you upgrade from 1.6 to 1.7:

If you already have added URLs with wildcards, Prisma Access might add URLs that duplicate existing URLs after the upgrade.

Prisma Access does not support FTP data transfers in active mode.

When Prisma Access performs a dataplane upgrade on a mobile user instance (an upgrade to a Prisma Access gateway or portal), any failed commits on the instance that were performed before the upgrade will not be applied to the upgraded instance.

During a Prisma Access dataplane upgrade, BGP statistics may not be available for 30 minutes in the Network Details page. This unavailability has no impact on dataplane traffic.

If you use Microsoft Edge or Firefox when using traffic steering, the browser does not forward traffic on its first attempt.

Workaround: Refresh the browser, then retry the operation.

If, in a traffic steering deployment with multiple traffic forwarding rules, two URLs in two separate rules resolve to the same IP address, Prisma Access sends traffic to the first rule in the list and will not use the second traffic rule. Traffic steering evaluates multiple traffic forwarding rules in order from top to bottom.

For a Prisma Access deployment with two Panoramas configured in high availability, you are able to request an upgrade to the GlobalProtect software version on the passive Panorama. Software upgrade requests are not applied if you request them on the passive Panorama.

Workaround: Do not request software upgrades on the passive Panorama; only request upgrades using the active Panorama.

When using traffic steering, Palo Alto Networks does not recommend using multiple service connections (whether dedicated or non-dedicated) in a target service connection group that is referenced in a traffic steering rule.

Prisma Access does not support a rule type of Intrazone if the source and destination zones are both Trust.

When entering CLI to retrieve Prisma Access job status, an invalid token message is received.

If you enable ECMP on a remote network, the values shown in the Statistics tab under PanoramaCloud ServicesStatusMonitorRemote Networks for Ingress Peak Bandwidth (Mbps) are correct; however, if you click the hyperlink for this value, the pop-up window that displays might show an incorrect value.

When creating a new mobile user deployment in multi-tenant mode, you receive an error that the Portal Hostname is not available when you assign it during mobile user onboarding.

Workaround: Before you begin your mobile user configuration, add an Infrastructure Subnet, commit all your changes to Panorama, and push the configuration changes to Prisma Access.

Some files are being skipped for DLP scanning when using OneDrive to upload multiple files.

When using DLP on Prisma Access, you can upload up to 25 files at a time.

When attaching a parent Device Group to a new remote network tenant in multi-tenant mode, the administrator is unable to attach device groups and templates.

Workaround: Log out, then log back in to Panorama.

If you use Box to upload multiple files, and one or more of the files are larger than 5 MB, the upload of all files will not complete. To continue, find the files in Box that are larger than 5 MB and click X to stop the download of those files.

DLP on Prisma Access is not supported in a Prisma Access multi-tenant deployment.

If you change the master key in Panorama (in DeviceMaster Key and Diagnostics), the master key for Cloud Services is not synchronized with this master key.

Workaround: Select PanoramaCloud ServicesConfigurationService SetupService OperationsEdit Master Key and manually change the master key to be the same as the Panorama master key.

When using Slack to upload multiple files, the Slack client treats the multiple file upload as a single request. If one of the files is not successfully uploaded, Slack retries the upload of all files a maximum of three times. If, after three retries, Slack cannot upload one or more of the files, the Slack client displays an error in the UI and doesn't upload any of the files.

When you upload a file using Slack, and the file is blocked, Slack detects the block operation as an upload failure and retries the file upload, which results in the same file being uploaded and blocked twice.

Workaround: This is normal Slack file upload behavior. Be aware that a single file that is uploaded using Slack might appear twice in the data filtering logs as being blocked.

When you delete a data filtering profile from a Prisma Access device group that is not shared, the profile name still appears when you add or configure a Security Profile Group, in the Data Filtering Profile area.

In a GlobalProtect deployment where the portal has multiple agent configs, when a GlobalProtect client logs in using the app, the portal looks for a matching agent config for the client by checking its OS type along with the config selection criteria. The agent configs are checked from top to bottom. If the OS type matches, but the config selection criteria does not, GlobalProtect marks the agent config as non-matching and moves to the next agent config to check for a match; however it no longer checks the OS type in these agent configs, and only looks for a match of the config selection criteria. This condition can cause the client to receive an agent config that has matching config selection criteria, but a non-matching OS type.

When configuring HIP redistribution, you cannot retrieve HIP information and set policies for the following use cases:

When using DLP on Prisma Access, when you upload a .docx file using SharePoint that was exported from Google Docs, the upload fails.

When setting up the GlobalProtect gateway connection settings (NetworkGlobalProtectGatewaysAgentConnection Settings) and specifying a Netmask to Restrict Authentication Cookie Usage, the commit fails if only a Source IPv4 Netmask is specified.

Workaround: Specify a Source IPv6 Netmask of 0, which disables the option for the specified IP address type.

If using Slack, Box, or Gmail to upload a file using DLP on Prisma Access, the response page is not displayed to the client if the upload is blocked.

Reverse DNS queries do not work in Prisma Access.

Workaround: Because type A and AAAA queries for internal domains work, you can specify *.in-addr.arpa in a query so that Prisma Access sends all reverse DNS queries to internal DNS servers.

When performing a Commit and Push operation for the Clean Pipe service, you receive an error that the Clean Pipe service had insufficient license resources, even though you have sufficient licensed bandwidth.

Workaround: Select PanoramaLicenses, then select Retrieve license keys from license server to retrieve the Clean Pipe licenses again.

If you add an existing template under one of the template stacks of Prisma Access (for example, Service_Conn_Template_Stack, Mobile_User_Template_Stack, or Remote_Network_Template_Stack), you cannot use objects of the added template in other Prisma Access templates that are part of the same template stack.

Previously, you could view and use objects from existing templates in Prisma Access templates if the templates were a part of a Prisma Access-specific template stack, which is not standard Panorama behavior.

In multi-tenant mode, Prisma Access automatically creates a set of templates, template stacks, and device groups for each tenant you create for remote networks, mobile users, and the Clean Pipe service. Prisma Access creates tenant-specific sets for all products, even if you are licensed for only one Prisma Access type.

When you delete a tenant, Prisma Access deletes the template and device group set for which you are licensed, but does not delete the unlicensed set. For example, if you have a remote network deployment and delete a tenant, Prisma Access does not delete the set it created for the mobile users and Clean Pipe.

Workaround: Manually delete the unused, unlicensed set of templates, template stacks, and device groups after you delete a tenant.

The Traffic Forwarding feature (PanoramaCloud ServicesConfigurationService SetupSettingsTraffic Forwarding) is not supported with multi-tenant deployments.

When you log out a Prisma Access mobile user from the Current Users window, the user still displays in the window after the logout operation.

Workaround: Close and then reopen the Current Users window to show the correct user status.

If you have two Panoramas set up in an active-primary and passive-secondary setup for Prisma Access, you cannot log out mobile users from the passive-secondary Panorama.

When you try to configure an Infrastructure Subnet (PanoramaCloud Services ConfigurationService SetupSettings) in multi-tenant mode, you can receive an Operation Failed message.

Workaround: Refresh the Panorama UI to have Prisma Access correctly apply the infrastructure subnet to the tenant's configuration.

When you perform a Commit All operation for mobile users, Prisma Access should display the commit status for portals and gateways separately; however, Prisma Access is displaying failures for portals under gateway status, and is displaying commit failures for gateways under portal status.

Workaround: Enter the debug plugins cloud_services prisma-access get-job-result jobid commit-job-id-number command, where commit-job-id-number is the ID of the commit operation that failed, to check and verify the commit operation for portals and gateways.

Pre-defined IKE Crypto, IPSec Crypto, and IKE Gateways templates do not display.

Workaround: Select PanoramaCloud ServicesConfigurationService Setup (for service connections) or PanoramaCloud ServicesConfigurationRemote Networks (for remote network connections), click the gear icon in the Settings area to open the Settings, then click OK.

When in multi-tenant mode, if you create a custom admin user with an Admin Role Profile that has Read Only access to the Panorama tab and has Plugin access disabled, that user can view, configure, and commit changes for subtenants.

Workaround: Disable access to the Panorama tab in the Admin Role Profile.

When you configure Clientless VPN with Prisma Access, the default security rule configuration uses the application-default service, which blocks clientless-vpn traffic.

Workaround: Change the default security rule to any service or service-http and service-https.

When configuring multi-tenant, if you create any device groups that are children or grandchildren of other device groups you create under the Shared parent device group, select only the device group at the lowest hierarchical level (child or grandchild) when you associate the device group to an access domain; do not select the parent.

You cannot reset the rule hit count for all Authentication and Application Override policies.

Workaround: Reset rules using a list of rules or a rule name for Authentication and Application Override policies.

When you migrate a single tenant to multi-tenant mode, you must do a local commit and then push the configuration before you add more tenants.

After upgrading to a new version of the Cloud Services plugin, you are able to downgrade. The downgrade operation should be disallowed.

Workaround: Do not downgrade the Cloud Services plugin after you have upgraded it.

When using the multi-tenant feature and migrating the first tenant to multi-tenancy, you can select template stacks and templates that are not associated with the tenant that you want to migrate, including templates that are used with on-premise firewalls.

Workaround: When you convert to multi-tenant mode, be sure to choose only those templates that you want to associate to the first tenant to migrate.

When configuring multi-tenancy, if you are planning to later configure Prisma Access for mobile users, you must do a local Commit of the your changes for the plugin (CommitCommit to Panorama) after you add templates, template stacks, and device groups for each tenant and before you onboard each tenant.

When using the multi-tenancy feature, users who manage single tenants cannot see the system logs. The MonitorLogsSystem choice is not available. This limitation applies to all Administrators who have an administrative role of Device Group and Template. Only superusers can view system logs in multi-tenancy mode.

When using the multi-tenancy feature and logged in as a tenant-level administrative user, opening the Panorama Task Manager (clicking Tasks at the bottom of the Panorama web interface) shows all tasks for all tenants, including any tasks done at the superuser (Admin) level.

When you enable multi-tenancy and migrate your configuration to the first sub-tenant, CLI commands are not supported for this operation. As a result, you must, use the Panorama user interface (UI).

If you configure a mobile user IP address pool for a single region instead of Worldwide, mobile users can still view and attempt to connect to all available gateway regions from their GlobalProtect app. This attempt fails because there is no IP address pool to allocate for other regions.

Workaround: To allow mobile users to manually select a gateway, either configure an IP address pool for the region in the location where you want the users to connect, or configure a Worldwide IP address pool for mobile users in Prisma Access to allow them to select all the locations you have deployed.

In an environment with on-premise firewalls on each side of Prisma Access and the remote network connections to which the on-premise firewalls are connected are in different regions, users behind one on-premise firewall cannot contact users behind another on-premise firewall unless you have configured an explicit policy to allow traffic between zone Trust and zone Trust.

If you change the master key in Panorama (in Device > Master Key and Diagnostics), the master key for Cloud Services is not synchronized with this master key.

Workaround: Select Panorama > Cloud Services > Configuration > Service Setup > Service Operations > Edit Master Key and manually change the master key to be the same as the Panorama master key.

When regular dynamic updates are downloaded to Panorama (by default, every Wednesday at 01:02), the MD5 checksum is changed. This condition can cause the Panorama configuration and the Prisma Access infrastructure to lose synchronization. While no tunnels are affected by this out of synchronization state, the status for Service Connections, Remote Networks, Mobile Users, and the Logging Service show a Config Status of Out of Sync.

Workaround: Perform a Commit and Push operation on the Panorama.

The BGP router configuration on the Prisma Access firewalls can receive a maximum of 15000 prefixes from each peer. And the total number of routes (static and dynamic) learned through BGP cannot exceed 25000. Exporting more than 25000 routes may adversely affect traffic flow on your network.

After you generate a new API key by selecting Panorama Cloud ServicesConfigurationService SetupGenerate new API Key, the previous API key is still valid for a period of time (up to five minutes). You use this API to retrieve the list of IP addresses for your Prisma Access firewalls.

For service and remote network connections that have BGP enabled, the Prisma Access ignores any route it receives from a neighbor with an AS number in its AS_PATH list that duplicates an AS number in the Prisma Access AS infrastructure (Infra-AS).

If you have configured a Notification URL, when you onboard a new remote network location, two notifications are sent to the URL instead of only one.

When you configure the same AS number for the service connection and remote network location(s), the routes are not imported in to the firewall on the remote network location.

Mobile users cannot connect to remote network locations without a service connection.

When configuring SAML, you must perform all configuration with a role of Superuser, including any configuration you perform for SAML using CLI.

The PanoramaCloud ServicesConfiguration page is grayed out when Panorama is not in sync with NTP.

Workaround: Make sure to synchronize time with NTP (PanoramaSetupServicesNTP).

Master Keys do not work for two Panorama appliances set as HA primary and secondary appliances.

Workaround: Deselect the Enable HA check box on the secondary Panorama appliance and commit the changes, set the same Master Key on both the primary and secondary Panorama appliance, then re-enable HA on the secondary Panorama appliance and commit the changes.

The DeviceSetupManagement page is not available on the Panorama appliance running the Prisma Access plugin. You cannot configure NT LAN Manager (NTLM).

Although Panorama allows you to delete the Mobile_User_Template that was created when the Prisma Access was provisioned, deleting this template also deletes your onboarding configuration and, upon commit, removes your Prisma Access for mobile users configuration.

When you onboard a new service connection or a remote network, the count for service connection and total remote peers displayed on PanoramaCloud ServicesStatusStatus is inaccurate until the provisioning is complete.

On Panorama, you cannot validate commit on a device group or template configuration before pushing the configuration to the Prisma Access infrastructure for remote networks and mobile users.