Grant Albitz wrote:
Local dd results:
write 268.343296 GB via dd, please wait...
time dd if=/dev/zero of=/PSC.Net/dd.tst bs=2048000 count=131027
131027+0 records in
131027+0 records out
real 3:09.8
user 0.1
sys 2:40.1
268.343296 GB in 189.8s = 1413.82 MB/s Write
131027+0 records in
131027+0 records out
268343296000 bytes (268 GB) copied, 146.209 s, 1.8 GB/s
real 2m26.215s
user 0m0.202s
sys 2m24.469s
so I am getting expected results locally. The network infrastructure is as
follows:
2 esxi hosts with intel x520 10g Ethernet cards. They are directly connected
via twinx cables to intel x520 network cards in the zfs box. So there is no
switch in between. Mtu is set on the zfs host as well as the esxi hosts to
9000. Results from both esxi hosts are the same.
Since I can get 700-800MBps write, you would assume the link could handle more
than 350MBps read...
There may be some vmware issue I am hitting, but I had the same setup before
with slower disks and I netted better performance in vmware. Are there any
settings in Comstar I can look at that could be causing this?
Have you tried the connecting to the volume form another IO host to
eliminate vmware as the cause?
--
Ian.
_______________________________________________
OpenIndiana-discuss mailing list
[email protected]
http://openindiana.org/mailman/listinfo/openindiana-discuss
