I thought I'd kick this forum off by asking what everyone uses ESX for?

We initially installed ESX onto 4 HP DL380 G3's with 12GB RAM each and all connected to a SAN.

These initally hosted two for Prodution network connected VMs and two for an isolated test lab that mirrored Prod. Then we got too many Dev/test servers and had to move one of the test lab DL380's to join the first two.

We're running 34 servers across 3 HP DL380's and 18 servers/workstations on the 4th DL380.

The first 3 run at around 30% CPU and the 4th runs at about 50-60% CPU.

We've got VirtualCenter 2.0.1 in and ready for migration to ESX3.

Moving to ESX3

We've put in two HP BL45p 4CPU dual core servers with 16GB RAM each. These will take over from the 4th DL380 and host our isolated test and a new Dev lab. the 4th DL380 will join in with the others and host production attached VMs.

These along with upgrades of the existing 2.5.3 servers to ESX 3.0.1 will be all up and running before christmas. (fingers crossed!)


We've also decided to go with Microsoft MSDN Pro licenses for all our Tech staff so that we don't need to purchase individual server licenses for all the dev/test servers.


What does everyone else's set ups look like?

Cheers,

Steve.

Hi and welcome!

We started with a number of GSX servers running dev and test. Soon we realised to take full advantage of virtualisation technology we should start using ESX. Originally kicked off as a better way to get into a realistic DR environment we bought a number of DL380's specced up with 8gb of ram and cheap SATA disk.

The DR environment pushed us forward with confidence in the VMware products and soon we were running it on HP DL380's in Prod and a HP Blade centre attached to a MSA1000.

More and more servers kept piling onto the virtual environment and soon we needed new infrastructure. So we bought a bunch of IBM blades attached to a couple of DS4800 SAN controllers with huge amounts of disk. The 2 bladecentres are populated with HS40's and HS20's and I am currently 80 percent of the way through my ESX 3 migration.

The ESX 3 side of things is now configured to run in HA and DRS mode and I must say it is awesome. It just opens up the next level in the VMware world for us, not having to continually monitor which host is getting hammered etc.

We also have set up a ESX 3 HA and DRS setup on another international site attached to our network. This site has it's own virutalised virtualcenter installation running on it. The licenses are floating internationally to a centralised license server though.

Anyone else ?

I'm in the middle of implimenting ESX 3.0.1 on three dual cpu dual core DL385's with an EVA4000 back end.

The syetm will virtualise around 24 servers, largely limited by the fact that the 385's will only scale to 16GB of RAM!

Doing DRS & HA although due to the 16GB limit have had to set the HA to allow violation of resource constraints.

Dev and Prod environments being spread across the servers & controlled by resource reservations.

Networking by trunking VLANs into ESX hosts using 8 x GBE in etherchannel & then port groups for each VLAN.

We are currently running a mixed environment of (3) HP DL580 G2 with 16GB RAM and (12) Dell 6850 with 32GB of RAM, supporting 217 vms. We are phasing out the HP servers since they are end of life.

We currently are running both ESX 2.5.x and 3.0.1, with a task of migrating everything to 3.0.1 by the end of June. Have not yet implemented DRS, HA, or consolidated backup.

We have moved almost all of our development and QA servers over to ESX.

We started from mid 2005 with a trial ESX2.x

Last July we built production on ESX3.0/VC2.0. Currently there are 3 NEC 140Rd-4, each with 2 or 4 CPUs and 24 GB of Memory. A fourth one is coming in next couple of weeks. SAN storage is NEC S2400. There are about 60 VMs running on this environment most of them are windows 2003 servers. The number is rising in a weekly basis.

DRS and HA have been turn on. We had a couple of occations that a VM had been rebooted after VMotions. We set DRS to semi-auto mode so that we can monitor a VM closely when we maually apply migration recommendations.

As part of DR plan, we do weekly VCB full VM backup to most of the VMs. The VCB backup process is great as it just takes significan loads off from ESX host CPUs and LAN. However, the restore process is timely and complcated. Hope the next version would improve.

We currently are in the process of network restructure and ESX3.01/VC2.01 upgrade