Hello xgrv,
thanks for your reply. You are Crazy ^^ i will test it with fio in the next days... I use a 3 Solaris ZFS Targets with IB... for 4 ESXi 5.5 Server. My parameter looks like
~ # esxcli system module parameters list -m ib_srp
Name Type Value Description
-------------------- ---- ----- ------------------------------------------------------------------------------
dead_state_time int 5 Number of minutes a target can be in DEAD state before moving to REMOVED state
debug_level int Set debug level (1)
heap_initial int Initial heap size allocated for the driver.
heap_max int Maximum attainable heap size for the driver.
max_srp_targets int Max number of srp targets per scsi host (ie. HCA)
max_vmhbas int 1 Maximum number of vmhba(s) per physical port (0<x<8)
mellanox_workarounds int Enable workarounds for Mellanox SRP target bugs if != 0
srp_can_queue int 1024 Max number of commands can queue per scsi_host ie. HCA
srp_cmd_per_lun int 64 Max number of commands can queue per lun
srp_sg_tablesize int 32 Max number of scatter lists supportted per IO - default is 32
topspin_workarounds int Enable workarounds for Topspin/Cisco SRP target bugs if != 0
use_fmr int Enable/disable FMR support (1)
What did you mean with ""H:0x5 D:0x0 P:0x0 error" was corrected by optimizing the latency" ???
I think i had a similar Problem with FC Cards in ESXi... That was the reason why we change to Infiniband and it is great. Hopefully we found a solution for this bad problem...
Thank you very much xgrv... If i can help you, i will do what i can...
Thanks
Thomas