LinuxKI Warnings - HewlettPackard/LinuxKI GitHub Wiki

LinuxKI Warnings are detected by the LinuxKI Kparse report (kiinfo -kparse) by looking at the trace records and identifying specific patterns. Each warning indicates a that a potential problem may exist on the system and further investigation is required to verify whether or not the warning is valid or if necessary action required to address the warning. Also, the indicated warnings may or may not actually contribute to a performance degradation on the system.

You can find more information on each of the warnings below:

• Benefits and Costs of KVM Halt Polling
• Heavy spinlock contention using mq-deadline IO scheduler with high IOPS
• Suspect SOS_BLOCKALLOCPARTIALLIST spinlock contention
• Performance degradation due incorrect clocksource
• High System CPU utilization during memory allocations, deallocations, and page faults
• ​​​​RunQ delays for critical processes can impact performance in a variety of ways
• Performance degradation on Microsoft Windows due to TCP interrupt timeouts
• Microsoft SQLServer scaling issues caused by SQL auto statistics
• Excessive page faults on KVM host
• Large IOs (>1MB) causing performance degradation on servers with PCIe Smart Array Controllers
• Oracle column tracking causing high CPU usage by Oracle processes
• Side-Channel Attack mitigation
• High SYS CPU time by processes reading /proc/stat such
• hugetlb_fault lock contention
• Excessive CPU time in pcc_cpufreq driver
• Excessive poll() calls by Oracle
• High wait time in md_flush()
• High BLOCK SoftIRQ times
• Network Latency Tuned profile
• Barrier Writes
• Power vs. Performance
• Unaligned Direct IO
• NUMA Balancing
• NUMA Off
• SAP DB2 semget
• Semaphore Lock Scaling
• Tasklet IRQs
• Unterminated ixgbe NICs
• Poor Direct IO Reads
• RHEL 7.3 Multipath round-robin bug