5.4 Upgrade Verification, Post-Upgrade Health Checks, & Troubleshooting
Key Takeaways
- Post-upgrade verification requires confirming Gaia OS version with fw ver -k, verifying installed hotfix takes with cpinfo -y all, and inspecting cluster synchronization with cphaprob stat.
- Security policy re-installation from the upgraded Security Management Server is mandatory after gateway major upgrades so object definitions and inspection binaries match R82 management capabilities.
- SmartConsole Central Deployment tasks and CDT CLI logs both provide per-target status; CDT writes detailed traces under /var/log/cdt/ on the SMS.
- Snapshot/rollback strategies use CPUSE revert (installer revert <package_id>) or Gaia OS snapshots (set snapshot revert <name>) when post-upgrade regression occurs.
Post-Upgrade Verification Commands & Health Verification
Once SmartConsole Central Deployment, a CDT batch plan, or a local CPUSE upgrade finishes, the CCSE administrator must execute mandatory verification and diagnostic checks on all target gateways before declaring the maintenance window successful and returning traffic to production.
1. Verification of Major Release & Kernel Version
To confirm that the Security Gateway kernel has successfully upgraded to Check Point R82 and is running the expected build number:
[Expert@Gateway1:0]# fw ver -k
This is Check Point VPN-1(TM) & FireWall-1(TM) R82 - Build 042 Kernel version 6.10
fw ver: Displays the major version and build number of Firewall-1 security software.fw ver -k: Displays the specific running kernel version, verifying that the kernel module loaded during boot corresponds to R82.
2. Detailed Hotfix & JHA Inventory (cpinfo -y all)
To list all installed hotfixes, Jumbo Hotfix Accumulator Takes, and minor packages currently active on the system:
[Expert@Gateway1:0]# cpinfo -y all
******************************************************************
HOTFIX INFORMATION:
HOTFIX_R82_JHF_MAIN Take: 25
HOTFIX_R82_PUBLIC_FIX_012 Take: 3
******************************************************************
cpinfo -y all queries the Gaia software registry and reports the exact JHA Take number, verifying that the intended hotfix build was applied.
3. ClusterXL Synchronization & State Verification
For cluster deployments, verify that ClusterXL membership, state synchronization, and interface monitoring are fully operational:
[Expert@Gateway1:0]# cphaprob stat
Cluster Mode: High Availability (Active-Attention / Standby)
Number Unique Address Assigned status
1 192.168.1.1 0 Active
2 (local) 192.168.1.2 1 Standby
[Expert@Gateway1:0]# cphaprob -a if
Required interfaces: 4
Required secured interfaces: 1
Interface Name: eth0 Status: UP
Interface Name: eth1 Status: UP
Interface Name: sync Status: UP (Secured)
Both members must show valid states (Active / Standby), and all cluster interfaces must report Status: UP with state synchronization active.
Security Policy Re-installation & Compilation Requirements
Following a major OS upgrade (e.g., from R81.20 to R82), Security Gateways cannot process new inspection features until a updated Security Policy is compiled and installed from the upgraded Security Management Server.
+---------------------------------------------------------------------------------+
| SECURITY MANAGEMENT SERVER (R82 SMS) |
| |
| 1. Update Gateway Object Properties -> Change Version to 'R82' |
| 2. Publish Database Changes in SmartConsole |
| 3. Install Security Policy (Compiles R82 Kernel Inspection Binaries) |
+---------------------------------------------------------------------------------+
|
Policy Download Tunnel (TCP 18191 / 18192)
v
+---------------------------------------------------------------------------------+
| UPGRADED SECURITY GATEWAY (R82 KERNEL) |
| |
| Receives updated inspection code ($FWDIR/state/local/FW1/ policy binaries) |
+---------------------------------------------------------------------------------+
Steps for Mandatory Post-Upgrade Policy Push:
- Open SmartConsole and navigate to
Gateways & Servers. - Open the properties of the upgraded Gateway / Cluster object.
- Change the Version drop-down menu from
R81.20toR82. - Click Publish to commit management database changes.
- Select Install Policy, choose target Access Control and Threat Prevention policy packages, and execute installation.
CDT Log Inspection & Debugging Framework
When a CDT deployment plan fails or encounters errors on specific target gateways, the administrator must inspect CDT's central log repository located on the Security Management Server at /var/log/cdt/.
Key CDT Log Files
cdt_trace_<plan_name>_<timestamp>.log: High-level execution timeline showing plan progression, phase starts/finishes, and overall gateway pass/fail status.cdt_debug_<plan_name>_<timestamp>.log: Detailed debug output capturing raw SSH transactions, CPUSE XML API exchanges, pre-check failure reasons, and script output./var/log/cdt/scripts/: Contains execution logs for custom scripts invoked by<execute_script>tags.
Common CDT Error Codes & Diagnostic Matrix
| Error Code / Log Pattern | Root Cause | Remediation Action |
|---|---|---|
ERR_PRECHECK_DISK_SPACE (101) | Target /var/log partition has less than 30 GB free space (R82 CPUSE package minimum) | Delete old CPUSE packages (installer delete local <id>) or clear crash dumps in /var/log/dump/ |
ERR_SIC_FAILED (102) | Trust certificate invalid or TCP 18191 blocked | Test SIC in SmartConsole; reset SIC trust via cpconfig if necessary |
ERR_CPUSE_BUILD_OLD (105) | Target CPUSE DA Agent build is below required minimum (1972) | Update CPUSE agent on gateway using installer update-agent |
ERR_SCRIPT_FAILED (108) | Custom shell script in <execute_script> exited with non-zero status | Inspect /var/log/cdt/scripts/ output log; fix script syntax or permissions |
ERR_TIMEOUT_REBOOT (112) | Gateway failed to rejoin network within expected timeout post-reboot | Connect via Lights Out Management (LOM/iLO) or serial console to inspect boot process |
Rollback Strategies & Procedures
If a post-upgrade health check reveals severe service degradation, application incompatibility, or kernel instability, administrators must execute a clean, controlled rollback.
1. CPUSE Package Revert (Hotfixes & JHAs)
For Jumbo Hotfix Accumulator updates, CPUSE maintains automated revert snapshots that allow instant uninstallation of the package without full OS re-imaging:
# View installed packages and identify package ID to revert
Gateway> installer show-installed
# Revert specific JHA package in Gaia Clish
Gateway> installer revert 1
CPUSE restores pre-installation binaries, updates system registry metadata, and prompts for gateway reboot.
2. Gaia OS Snapshot Rollback (Major Release Upgrades)
For major OS upgrades, reverting via CPUSE may not be supported. Administrators rely on Gaia OS Snapshots created prior to upgrade initiation:
# Display available OS snapshots stored on local disk
Gateway> show snapshots
# Revert system to pre-upgrade snapshot image
Gateway> set snapshot revert Pre_R82_Upgrade_Snapshot
Comprehensive Rollback Method Comparison
| Rollback Mechanism | Target Scope | Recovery Duration | Data / Config Loss Risk |
|---|---|---|---|
| CPUSE Package Revert | Specific JHA or hotfix package | 5 to 10 minutes (requires reboot) | Minimal; uninstalls hotfix binaries while retaining configuration |
| Gaia OS Snapshot Revert | Entire OS image (Kernel, Network, Local Config) | 15 to 25 minutes | Reverts entire local disk image to snapshot creation timestamp |
| Gaia OS Backup Restore | Configuration database & network settings | 10 to 15 minutes | Restores system settings; requires base OS of matching version |
| LOM / USB Clean Re-image | Total hardware appliance bare-metal restore | 30 to 60 minutes | High; completely wipes appliance drive; requires full reconfiguration |
Which command provides a comprehensive breakdown of all installed Check Point hotfixes, Jumbo Hotfix Accumulator takes, and CPUSE packages on a Security Gateway?
Where are CDT execution trace logs and debug logs centrally stored on the Security Management Server during a deployment plan execution?
If a Security Gateway experiences severe traffic handling issues immediately after installing a new Jumbo Hotfix Accumulator via CPUSE, what is the fastest CLI method to safely restore the previous hotfix state?
What is a mandatory post-upgrade step required on SmartConsole after successfully completing a major Gaia OS gateway upgrade from R81.20 to R82?