Came across following bug while carrying out assessment for N1000v. Has anyone came across this yet? This seem to be a serious bug.
Symptom:
ESXi host becomes unreachable because of a PSOD in sf_qos_list_set(). The entire backtrace looks like the following:
sf_qos_list_set@com.vmware.vmkapi#
sf_dp_qos_lookup@com.vmware.vmkapi#
sf_scheduler_exit@com.vmware.vmkapi#
sf_sched_packet_list@com.vmware.vmkapi#
svs_packet_list_receive@com.vmware.vmkapi#
svs_set_pcpu_pshandle@com.vmware.vmkapi#
Port_InputResume@vmkernel#
Port_Input_Committed@vmkernel#
Vmxnet3VMKDevTQDoTx@vmkernel#
Vmxnet3VMKDev_AsyncTx@vmkernel#
NetWorldletPerVMCB@vmkernel#
WorldletBHHandler@vmkernel#
BH_Check@vmkernel#
CpuSchedIdleLoopInt@vmkernel#
CpuSched_IdleLoop@vmkernel#
Conditions:
This PSOD happens on Nexus 1000V release 5.2(1)SV3(1.15). QoS needs to be enabled.
Workaround:
Disable QoS
| Known Affected Releases: |
(1)
|
