Hi folks,
this is my very first posting in this community and I am really looking forward go get help 🙂
I am using 3PAR F400 as storage for our C-Class Blade Enclosure hosting 4x BL460C G6 blades. Everything's in place and working properly so far. The 4 blades have access to the 2x2TB luns created for VMware. The virtual machines are working very well.
To furhter improve performance, we created LUNs and mounted them as RAW devices straight to the virtual machines (so no virtual disk). But instead getting better performance, the performance is poor! Well it depends, sometimes it's better and sometimes worse but always insufficient to run our SQL servers.
I found the following article which applies to our environment. I have not yet tested it and need to wait for the 5th blade to arrive for testing. Have you experienced simiar issues???
"Controlling LUN queue depth throttling in VMware ESX/ESXi (KB 1008113)
Thank you in advance!
Hello and welcome to the forums.
To furhter improve performance, we created LUNs and mounted them as RAW devices straight to the virtual machines (so no virtual disk). But instead getting better performance, the performance is poor! Well it depends, sometimes it's better and sometimes worse but always insufficient to run our SQL servers.
You won't gain performance improvements with RDM. The differences are negligible these days.
How many VMs are you running on these 2 datastores total? It sounds like you probably have latency issues, which are showing up intermittently on the RDMs.
Good Luck!
Hello and many thanks for your quick answer!
We're running approx. 80 virtual machines on these 2 LUNs. The RDM are presented to specific virtual machines only. I have noticed that performance is great on VMware disks while the performance on RDM is no good. But it should be similar anyway so there's a problem still, right?