Hi Everyone
I have an issue and I ran out of ideas so I need some help
SQl restore of a 1 GB file takes around 30 minutes and after a while I noticed Read/Write to the storage seems to be slow
Here my environment:
Windows 2012 Standard R2 x64 (virtual server running in esx)
Storage connected with iSCSi going thru the production network (I know is not recomended :S)
Test using DiskSpd.exe -c20G -d120 -r -w40 -t8 -o32 -b8K -h -L G:\temp\sqlIOTest.dat
Read test using diskspd.exe
Write IO , AvgLat = 323.391
Read IO , AvgLat = 323.003
Versus another server (virtual server running in the same esx) in the same vlan, also same OS version, same configuration in regards of virtual hardware and Storage connected with iSCSi going thru the production network
Test using DiskSpd.exe -c20G -d120 -r -w40 -t8 -o32 -b8K -h -L G:\temp\sqlIOTest.dat
Write IO , AvgLat = 20.900
Read IO , AvgLat = 21.266
What could be causing this?
I know the iSCSi should be using an exclusive network but there is another cluster with the same conf and no issues
I already disabled TCP values (Chimeny, RSS, Netdma)
No performance issues, CPU and Memory are barely used at the time the restore is running
Any help would be highly appreciated
Best regards
Joaquin Camarero Muñoz