Environment: 2 VMWare VSphere 4 ESXi Hosts each have 2x 6-core CPUs and 48GB RAM
VMW hosts connected through 4GB fiber to FC Disk on EMX CSX3 Clariion.
Problematic Guest: Exch 2010 on Windows 2K8R2 VMDK lives in storage group A (104/134GB consumed)
Exchange vitrual disk lives in storage group B (250GB/270GB consumed)
On Saturday we began migrating mailboxes from Exchange 2003 on a physical machine to a virtual Exchange 2010 on VSphere 4. Just before the migration, we expanded the amount of RAM on the virtual Exchange 2010 server from 4GB RAM to 12GB of RAM.
Started the migration and the CPU pegged and memory was consumed. VMotioned all other VMs off the guest hosting the Exch2010 and increased RAM to 20GB and CPUs to 4. Still the Exch server became unresponsive.
Colleagues are pointing at disk IO not being able to handle the load.
My backup Exchange server which is hosted on identical storage and ESX environment runs smoothly with 4GB Ram.
Any ideas on what is causing my prod Exch server to crawl at snails pace? I noticed that by adding more RAM on the prod environment, the VMSW file has aldo grown to 20GB, but the Exch server is the only guest on a host with 48GB Ram.