SONiC Bug: Missing Kernel Modules On Quanta IX8-56X

by Alex Johnson 52 views

Introduction: The Critical Absence of Kernel Modules

SONiC (Software for Open Networking in the Cloud) is a crucial open-source network operating system, and the Quanta IX8-56X platform is a popular choice for its deployment. However, a significant issue has surfaced in the SONiC 202505 Broadcom image. The sonic-platform-quanta-ix8-56x_1.0_amd64.deb package, vital for platform-specific functionalities, is missing essential kernel modules. These modules, located in /lib/modules/$(uname -r)/extra/, are fundamental for the proper operation of the hardware. Their absence leads to severe operational deficiencies. This article delves into the specifics of this bug, its impact, and how it can be reproduced. The missing modules include crucial drivers like qci_platform_ix8, qci_cpld_sfp28, qci_cpld_led, and quanta_hwmon_ipmi. Without these, key hardware components and functionalities are rendered inoperable, affecting the overall performance and reliability of the network switch.

Detailed Description of the Bug

The core of the problem lies within the packaging of the SONiC image. The sonic-platform-quanta-ix8-56x_1.0_amd64.deb package should ideally contain all necessary kernel modules to interface with the platform's hardware components. The absence of these modules creates a cascading effect of failures. The I2C buses, essential for communication with SFP/EEPROM devices, are unavailable. Consequently, the system fails to recognize and manage connected SFP modules. The ix8-platform-init.service, designed to initialize the platform, fails to load the required modules, leading to error messages such as "modprobe: FATAL: Module qci_cpld_sfp28 not found". This service failure further cripples the system, preventing proper hardware initialization. Without the correct kernel modules, critical hardware functionalities, including SFP, sensor, PSU, fan, LED, and EEPROM access, become unavailable. The system health daemon crashes repeatedly because the sonic_platform package is not correctly installed. This package is essential for monitoring and managing the hardware's health and functionality. The bug affects core switch operations. This bug is likely a critical issue in the 202505 build pipeline.

Steps to Reproduce the Issue

Reproducing this bug is straightforward. First, install or boot the 202505 Broadcom image on a Quanta IX8-56X switch. Next, check the installed platform package to confirm its contents. Use the command dpkg -L sonic-platform-quanta-ix8-56x to list the files in the package. Subsequently, inspect the /lib/modules/6.1.0-29-2-amd64/extra/ directory to verify the absence of the essential kernel modules. To confirm the failure, examine the logs of the ix8-platform-init.service using sudo journalctl -u ix8-platform-init.service -b. The logs will show that the service fails to load the required modules. The system health daemon will also fail. Finally, verify the output of show version and show techsupport commands to provide system information for troubleshooting. The absence of kernel modules and the failure of platform initialization services will confirm the presence of the bug, emphasizing its critical impact on the platform's functionality.

Expected vs. Actual Behavior

The expected behavior is for the sonic-platform-quanta-ix8-56x_1.0_amd64.deb package to contain all required kernel modules under /lib/modules/$(uname -r)/extra/. This includes modules like qci_platform_ix8.ko, qci_cpld_sfp28.ko, qci_cpld_led.ko, and quanta_hwmon_ipmi.ko. This configuration ensures that all hardware components function correctly, and the system can monitor and manage the hardware. The package should function as seen in older SONiC images, such as the Quanta SONiC 25.06, which included the necessary modules for kernel 5.10. The actual behavior, however, deviates significantly. The package contains only sonic_platform-1.0-py3-none-any.whl and a service file, lacking all the essential kernel modules. The result is a non-functional system. The missing modules lead to critical operational failures, including missing I2C buses, inaccessible EEPROM, broken SFP/QSFP detection, non-functional LEDs, and missing FAN/PSU sensors. The system-health and watchdog processes fail repeatedly, leading to instability and making the platform unusable.

Impact and Severity

The impact of this bug is significant. It renders the Quanta IX8-56X platform partially or completely non-functional. The inability to access essential hardware components and the failure of platform initialization services severely limit the platform's ability to operate as a network switch. This directly affects the network's performance and stability. Because of these critical failures, the severity of the bug is high. The failure to include the required kernel modules prevents the platform from functioning correctly. This bug is a blocking issue for using the SONiC 202505 image on Quanta IX8-56X hardware. It should be addressed immediately to ensure the platform's reliability and usability.

Conclusion: Addressing the Packaging Defect

The core issue is a packaging bug in the SONiC 202505 build pipeline, specifically the omission of essential kernel modules from the platform package. To resolve this, the build process must be reviewed and corrected to ensure the inclusion of all necessary kernel modules in the sonic-platform-quanta-ix8-56x_1.0_amd64.deb package. This fix is crucial to restore the platform's full functionality and allow it to operate as intended. The bug's high severity and impact on the platform's basic operations highlight the importance of prompt action to resolve this issue and prevent further disruptions.

For more information on SONiC and its functionalities, you can visit the official SONiC documentation SONiC Documentation.