VxVM vxio V-5-3-1285 voldmp_errbuf_sio_start: Failed to flush the error buffer *** on device *** to DMP
In this VMware scenario, dmpnodes are actually vmdk disks, there is only one path per dmpnode. When the disks are removed, triggering the udev event,
DMP disables the only path of these dmpnodes. After that, DMP will do the I/O error analysis by calling dmp_flush_errbuf() which in turn calls dmp_process_errbuf().
In dmp_process_errbuf() we return error if the number of active paths of dmpnode is zero. Hence we see this warning messages.
This is current Vxdmp design and SFHA limitation on VMware environment. Please ignore this vxio messages when ServiceGroup switch or failover.
It won’t cause any IO losing.
Applies To
VCS on VMware environment
VMDK disk setting