rodfantana
Dabbler
- Joined
- Jun 10, 2017
- Messages
- 27
Hey guys hope all is well!
I'm running a virtualized FN 9.10-U3 on ESXi 6.5. Everything is set up according to generally accepted best practices as far as reservations go from the hypervisor perspective (CPU shares = high, memory is reserved and shares are high.)
Spec wise -
VM Specs: 2vCPUs (xeon D1541), 8GB RAM, 8GB vmdk that host FN and system vol.
RAIDZ1 data vol: LSI 9207-8i (in passthrough mode) with 4 x 6TB Toshiba X300 7200RPM drives with ECR set to 7sec at boot.
PROBLEM: When I try to read or write via CIFS mount (haven't tried other methods) I get about 90-100MB/sec from this data vol. That's the speed I get when i read or write a 30GB file; nother else is using that storage at the time. It also jumps between those numbers a lot. Other storage that sits on the same switch does consistent 112MB/sec w/o hickups and saturates a 1gps switch during the same file read/write. From looking at Reports in the GUI, I don't see anything that looks pegged, neither from ESX side.
Does anyone have any suggestions on what else to look at?
Thanks in advance.
~Rod.
I'm running a virtualized FN 9.10-U3 on ESXi 6.5. Everything is set up according to generally accepted best practices as far as reservations go from the hypervisor perspective (CPU shares = high, memory is reserved and shares are high.)
Spec wise -
VM Specs: 2vCPUs (xeon D1541), 8GB RAM, 8GB vmdk that host FN and system vol.
RAIDZ1 data vol: LSI 9207-8i (in passthrough mode) with 4 x 6TB Toshiba X300 7200RPM drives with ECR set to 7sec at boot.
PROBLEM: When I try to read or write via CIFS mount (haven't tried other methods) I get about 90-100MB/sec from this data vol. That's the speed I get when i read or write a 30GB file; nother else is using that storage at the time. It also jumps between those numbers a lot. Other storage that sits on the same switch does consistent 112MB/sec w/o hickups and saturates a 1gps switch during the same file read/write. From looking at Reports in the GUI, I don't see anything that looks pegged, neither from ESX side.
Does anyone have any suggestions on what else to look at?
Thanks in advance.
~Rod.