Prisma SD-WAN
Best Practices for Memory Efficient Security Policies on ION Devices
Table of Contents
Expand All
|
Collapse All
Prisma SD-WAN Docs
-
-
- Prisma SD-WAN Controller
-
- CloudBlade Integrations
- CloudBlades Integration with Prisma Access
-
-
-
-
- 6.5
- 6.4
- 6.3
- 6.1
- 5.6
- Prisma SD-WAN Controller
- Prisma SD-WAN On-Premises Controller
- Prisma SD-WAN CloudBlades
- Prisma Access CloudBlade Cloud Managed
- Prisma Access CloudBlade Panorama Managed
Best Practices for Memory Efficient Security Policies on ION Devices
Learn about the best practices for memory efficient security policies on ION
devices.
| Where Can I Use This? | What Do I Need? |
|---|---|
|
|
Overview
Prisma® SD-WAN software versions 6.3.6, 6.4.3, 6.5.3, and later introduce enhanced
features and capabilities. While these innovations provide significant value, they
can also increase memory utilization, particularly on ION models with limited memory
capacity such as ION 1000, ION 1200, and ION 2000.
In environments with extensive custom application definitions, large
prefix lists or large number of security zones referenced within security policies,
memory-intensive compilation processes may lead to resource exhaustion. This can
result in instability or unexpected device behavior, including device reboots or
loss of connectivity to Strata Cloud Manager, potentially requiring on-site
recovery.
The following sections outline best practices to help optimize memory usage and
maintain platform stability.
Understanding Memory Risk
Deployments that include the following configurations experience the most memory
pressure:
- Large custom application definitions with extensive prefix lists
- Large number of security zones
- Security policies containing numerous or highly granular prefixes (for example, many /32 entries)
- Frequent updates to prefix lists or policies
- Devices with limited physical memory (for example, ION 1000, ION 2000, and ION 1200)
Pre-Change Assessment
Before you implement security policy related configuration changes, complete the
following steps:
- Assess current memory utilization to determine risk exposure.
- Validate changes in a lab or sandbox environment prior to production deployment.
- Apply changes incrementally to minimize impact and isolate issues.
- Schedule updates during maintenance windows to reduce service disruption when possible.
Checking Current Memory Utilization
You can verify device memory status using Strata
Copilot.
Use natural language queries to identify at-risk devices. For example:
"Provide a table of all connected Prisma SD-WAN ION devices, including average and
maximum memory utilization, sorted by average memory utilization (high to low).
Ensure the table includes the ION model for each device."
Apply model filters (ION 1000, ION 2000, ION 1200) to focus on devices with limited
memory constraints.
In Strata Cloud Manager navigate to, Insights > ION
Devices > Device Activity to review historical memory utilization
graphs. Look for sustained high utilization patterns or increasing trends over
time.
Custom Application and Prefix Management
Efficient management of custom applications and prefix filters is critical for memory
optimization.
- Removing Unused Custom ApplicationsStale or unused custom application definitions consume memory during policy compilation even when no traffic matches them.
- Prefix AggregationReplace multiple smaller prefixes with larger subnets where appropriate. For example, consolidating 256 individual /32 addresses into a single /24 subnet reduces policy table entries by approximately 99% while maintaining equivalent coverage for homogeneous subnets.
- Managing /32 Prefix UsageExcessive use of /32 prefixes, especially in applications using broad port ranges, creates large policy tables during compilation. Each unique /32 entry multiplies the number of policy table entries when you combine it with source zones, destination zones, and port ranges.
- Port Range OptimizationUsing full port ranges (1-65535) in application definitions or security policies creates 65,535 individual port entries in the policy table. This multiplies memory requirements by several orders of magnitude compared to specifying discrete ports.
Security Policy Design Considerations
Security policy structure plays a significant role in memory consumption.
The most critical factor is the combination of broad zone assignments with large
prefix lists.
Avoid 'Any source zone' with 'Any destination zone' especially with large prefix
lists. This combination creates a large number of policy objects that significantly
increase memory usage, which can be catastrophic in a system with limited system
memory.
Important Considerations and Trade-offs
While you optimize for memory, be aware of the following impacts on operational
capabilities.
- Application Visibility Without Policy ReferencesWhen you remove custom application definitions (AppDefs) entirely, you eliminate both policy control and visibility. However, you can maintain visibility while reducing memory pressure by decoupling AppDefs from security policy references.
- Prefix Aggregation Impact on Policy MatchingConverting granular prefixes (for example, /32 to /24) may cause unintended policy matches if the aggregated subnet contains IPs with different policy requirements. Before you aggregate, validate that all IPs within the target subnet legitimately require the same policy treatment.
- Traffic Steering and Policy Enforcement Without AppDefsWithout custom application definitions you reference in policies, the system cannot steer application-specific traffic based on application characteristics. The system will base path selection decisions on IP prefixes, destination zones, and general traffic types rather than application-aware intelligence.
Recovery Process
When the ION is unable to compile and install large security policy updates due to
memory constraints, this may lead to service instability, including crashes,
unexpected reboots, and in some cases, loss of connectivity to the SD-WAN
Controller. The following recovery steps can be performed as the case may be:
- Reapply Security PolicyIn Strata Cloud Manager, unbind the Security Policy associated with the site. Wait approximately 30 seconds, then reapply the Security Policy to the site. Verify whether the ION returns to an online and stable state.
- Reboot the DeviceIf the issue persists and the ION remains offline, perform a reboot of the device and check the status again.
- Recover via Console AccessIf the reboot does not resolve the issue, the ION may be stuck in a continuous reboot cycle. In this case, console access is required to recover the device:
- Unbind the newly applied Security Policy from the site in Strata Cloud Manager.
- Connect a console cable to the ION and open a terminal application to access the device CLI. Configure the terminal settings to 115200/8/N/1.
- Execute the following command to stop the process:debug process stop name=fcThis step may take approximately 5 minutes.
- Verify if the ION has re-established connectivity with the SD-WAN Controller.
- Restart the process using the command:
debug process start name=fc- Reapply the last known working Security Policy to the site after the device is stable.