LUN Sizes and Raid Groups for VMWare

Question

Could someone direct me to some white papers that would help me determine what RAID groups and LUN sizes I should consider for a new VMware implementation.

We have a CX320 and looking to have 150 VM's converted over the next 3 years. I now need to determine what the best storage configuration would be to support this

I understand that RAID 10 would be the best RAID group to use from a performance perspective, but how can i determine the number of disk needed and the number of VM's it could support and if RAID 5 workable solution also.

Some non-EMC documents have recommended that is best to have many smaller LUN holding no more that 15-20 VM per LUN. but I¿m having difficulties finding information to quantify these numbers.

Thanks in advance.

Loudenj · Answer

Hi Reg,I would suggest in powerlink search for ESX, if you only have Documentation and Whitepapers selected it comes up with a good range of emc doco. Where I startedOur config is a single SG for the VM cluster, Luns of 300gb size for production VMFS (4gb FC Drives), 300gb for Dev	est (SataII), 2 300gb Luns for DR (SataII)Hosts on luns (average)Prod = 7Dev = 13DRP = 4OUr File server and change servers have RDM drives assigned to them. All production databases are physical box's)Currently we have 90 odd hosts, the dev ones get at thrasting as they contain sql dev db's. Disk is never a problem, CPU and Memory are.Hope this helps.John

mrcreosote · Answer

Hi Reg.Those recommendations aren't based on any limitations of the storage (regardless of vendor or disk config) - they are based on a current limitation of ESX.There is one SCSI command queue per vmfs datastore, therefore it is feasible that if you create a datastore large enough to hold a large number of VMs (or a smaller number of VMs with high IO profile) the command queue length will increase at the ESX host and become a bottle neck.IMHO, aim to have 8 - 12 VMs per datastore - normally means 250-300GB LUNS. Avoid adding extents at all costs.With regard to Raid levels, this depends entirely on what the VMs are running..... Try these whitepapers for a general read:http://www.vmware.com/pdf/vi3_san_design_deploy.pdfhttp://www.vmware.com/pdf/vi3_35/esx_3/r35u2/vi3_35_25_u2_san_cfg.pdf

kelleg · Answer

There are a couple of documents on PowerLink that will be helpful in gaining an understanding of the differences between the different Raid types support on CLARiiON.

EMC CLARiiON Best Practices for Fibre Channel Storage: FLARE Release 26 Firmware Update - Best Practices Planning

http://powerlink.emc.com/km/live1/en_US/Offering_Technical/White_Paper/H2358_clariion_best_prac_fibre_chnl_wp_ldv.pdf

EMC CLARiiON Fibre Channel Storage Fundamentals - Technology Concepts and Business Considerations

http://powerlink.emc.com/km/live1/en_US/Offering_Technical/White_Paper/H1049_emc_clariion_fibre_channel_storage_fundamentals_ldv.pdf

White Paper: The Influence of Priorities on CLARiiON LUN Management Operations - Applied Technology

http://powerlink.emc.com/km/live1/en_US/Offering_Technical/White_Paper/H4153-the-influence-of-priorities-on-lun-mgmt-optns-wp.pdf

MetaLUNs

http://powerlink.emc.com/km/live1/en_US/Offering_Technical/White_Paper/H1024.1_clariion_metaluns_cncpt_wp_ldv.pdf

Glen

hersh1 · Answer

3 years ago we started with 300 GB LUNs. We have recently made the switch to 500 GB and now 1 TB LUNs. The 1 TB LUNs actually hold less VMs per LUN then the 500 so far because of the large datastores that reside there. I think eventually that will increase since we just made the move to 1 TB in the past month.

CLARiiON

LUN Sizes and Raid Groups for VMWare

Was this post helpful?