Description of problem: BZ 1418065 describes the problem where, in a RHEV environment, when hosts are upgraded to RHEL 7.3 and VMs access a direct lun configured as virtio-scsi with scsi-passthrough enabled and the lun is accessed via an Emulex FC HBA, the VMs will be paused due to an i/o error. However, the problem is not in RHEV/VDSM, but in the Emulex driver. The current resolution to the bug is to increase the 'lpfc_sg_seg_cnt' in 'lpfc.conf'. Is it possible/feasible that VDSM can manage this configuration ? Version-Release number of selected component (if applicable): RHEV 3.6 RHV 4.0 How reproducible: 100% with the default 'lpfc_sg_seg_cnt' parameter. Steps to Reproduce: Refer to BZ 1418065. Actual results: Expected results: Additional info:
Moving to the Node team, ss the discussed solution should come from the platform and node should respin on top of it.
I hate the idea of owning another platform config file; I hope this can be solved below RHV (in the driver) or above RHV (by a customer deployment script). https://bugzilla.redhat.com/show_bug.cgi?id=1420473#c34
Once we
Test version: redhat-virtualization-host-4.1-20170417 imgbased-0.9.23-0.1.el7ev.noarch kernel-3.10.0-514.16.1.el7.x86_64 # rpm -qa| grep qemu-kvm qemu-kvm-rhev-2.6.0-28.el7_3.9.x86_64 qemu-kvm-tools-rhev-2.6.0-28.el7_3.9.x86_64 qemu-kvm-common-rhev-2.6.0-28.el7_3.9.x86_64 RHVH include the correct qemu-kvm-rhev version which already included the fix for bug 1431149. So the bug is fixed, change bug status to VERIFIED.
sync2jira