By Aaron Delp
Welcome to Part 4 in our series on blades and virtualization. Please note that I will be covering the HP virtualized I/O offerings in a near future post. I plan to cover Virtual Connect, Flex-10, and the BL495 blade at that time.
As always, let’s start by reviewing the hardware specification we will be using to compare the rack servers against the blade servers. Here are the portions of the spec that concern us:
- A minimum of 32 GB memory – As many have commented and my own studies have shown, memory is usually the first limiting factor. The more memory you can install in the system, the better.
- 2 on-board NICs + some combination of expansion cards (FC, iSCSI, NIC, 10Gb)
As we did for the IBM side, let’s start with memory. The maximum amount of memory available on the DL360 is currently 64GB when using 8GB DIMMS in each of the eight slots. The BL460c has the same memory configuration (8 slots), but has a few caveats depending on the processor power consumption. According to the HP Product Bulletin Tool, if the processor is 80W or below, then the max amount of memory is 64GB (8x8GB). If you would like to use processors above 80W, then the maximum amount of memory is 48GB in a 6x8GB configuration. Which processors take above 80W you ask? As of this writing, there are only 2 processors in this category, the 3.0GHz Quad Core (Intel X5450 chip, not the E5450 chip!) and the 3.16GHz Quad Core (Intel X5460).
As with the IBM solution, I will explore four possible transport technologies for virtualization: 10Gb, iSCSI, FC, and NFS. I will also fill any remaining expansion slots with 1Gb Ethernet ports. Just like the IBM Chassis, we have a maximum of eight expansion ports.
HP expansion is a little more straightforward than IBM. Bays 1 & 2 connect to the blade on-board NICs and must be populated with an Ethernet compatible switch. An HP BL460c blade has two expansion slots, labeled Mezzanine 1 & 2. Bays 3 & 4 on the chassis connect to the adapter in Mezzanine Slot 1 on the BL460c Blade. Bays 5-8 connect to Mezzanine bay 2 on the blade. If a dual port card is placed in Mezzanine 2, only Bays 5 & 6 will be active. A four port card is required (Ethernet is the only 4 port card currently) to access all four switch bays.
The only exception to the above port mappings is 10Gb. Like the IBM 10Gb switch, the HP 10Gb switch is a “double wide” switch. For dual 10Gb in the HP chassis, one switch must be placed in Bays 5/6, the other in Bays 7/8.
Here are the possible configurations given the port mappings above:
BL460c blade with 10Gb & 4 NICs:
| Dual On board NICs | Bays 1 & 2 (Ethernet switches) |
| Mezz 1 – Dual Port NIC | Bays 3 & 4 (Ethernet Switches) |
| Mezz 2 – Dual Port 10Gb Card | 5/6 – 10Gb switch & 7/8 10Gb switch |
BL460c blade with 10Gb, 2 NICs, and 2 FC:
| Dual On board NICs | Bays 1 & 2 (Ethernet switches) |
| Mezz 1 – Dual Port FC Card | Bays 3 & 4 (FC Switches) |
| Mezz 2 – Dual Port 10Gb Card | 5/6 – 10Gb switch & 7/8 10Gb switch |
BL460c blade with FC & 6 NICs:
| Dual On board NICs | Bays 1 & 2 (Ethernet switches) |
| Mezz 1 – Dual Port FC Card | Bays 3 & 4 (FC Switches) |
| Mezz 2 – Quad Port 1Gb Card | Bays 5-8 Ethernet switches |
BL460c blade with iSCSI & 6 NICs:
| Dual On board NICs | Bays 1 & 2 (Ethernet switches) |
| Mezz 1 – Dual Port iSCSI Card | Bays 3 & 4 (Ethernet Switches) |
| Mezz 2 – Quad Port 1Gb Card | Bays 5-8 Ethernet switches |
BL460c blade 8 NICs:
| Dual On board NICs | Bays 1 & 2 (Ethernet switches) |
| Mezz 1 – Dual Port 1Gb Card | Bays 3 & 4 (Ethernet Switches) |
| Mezz 2 – Quad Port 1Gb Card | Bays 5-8 Ethernet switches |
As you can see from the tables above, HP has a nice round portfolio without any holes. In the next section we will be covering the IBM virtualized I/O solutions.
Tags: bladevrack, Hardware, HP, IBM, Virtualization
-
Hi Aaron,
I’ve done a similar post on Blades and some advice on adoption, i’ve tried to steer clear of the excellent work you are doing and just stick to common sense arguements.
Any reason you havent done a FC and IP Storage option?
-
Hi Aaron,
it would be great to include also the VM otimized Blades like the BL495 with Flex10. Personally I think that this is THE VM base that is currently available on the market (esp. with the Flex10 because it give you an unmatched NIC flexibility.
And – No I dont get actually paid by HP
MAx
-
Ok … Probably I should read the article before I start to comment
-
Scott, a couple of things.
- Make sure that you are using the G1 version of the BL460, the G5 version supposedly only supports 32 GB of LP memory per HP Product Bulletin. I’m checking with HP about this.
- Really a better idea is just not to use the G5 version at all yet, it only supports 2 processor speeds, L5450 and the E5430. It is kinda funny that the PB states that 64GB kits are an option for the G5 though.
Have you noticed many companies using BL460 for their virtualization options? Seems kinda hard to get a decent ROI without going to the 480 or 680.
-
@Daniel – Do you mean a solution that is 2 nics, 2 FC, and 2 iSCSI? It would be possible to set it up using one each adapter but I see both FC and iSCSI as the transport protocols to the storage. In the ESX case, with only 2 nics left for VMotion, Service Console, and Virtual Machine traffic I believe that would be tight. You could do it but I’m afraid the NICS would get saturated and I have always broken up the traffic to achieve better utilization and segregation of the traffic. I would love to hear your thoughts on the subject!
-
Richard,
I actually do have a few sizable customers who are using BL460c blades as their virtualization platform. It really just depends upon your configuration–as Aaron has pointed out above, it’s possible to get 6 NICs and dual FC connections on a BL460c, which is fine for most virtualization implementations. You just have to size them right.
Thanks!
-
Guys,
I am recommending BL480/680 purely due to the limit in ports on day one for ESX if you want to deploy a shared DRS cluster with different tiers of storage.
If you deploy 6 Ethernet Modules and 2 FC Fabric modules you can according to the HP Sizing tool have 6 nics (2 onboard and 4 on mezzanine) per BL460 ESX Host….
-2 x Vmotion and SC (Using Port Groups) in Active/Active
-2 x Prod Network
-2 x Spare -
Daniel, how many implementations have you seen where the VMs actually generated close to or more than 2Gbps of network traffic? I’m curious, because 6 NICs for a VMware implementation with FC storage is, in my mind, plenty. (And that’s accounting for future needs, like FT logging NICs for that future product that we can’t discuss publicly!)
-
Depends how big your ratio is but your right, network is always the heaviest underutilised IO, specially when your P2V’ing 10 and 100MB older servers!
I like to have a bit of breathing space to be honest
And yes I think spare adapters is going to become usefull for the likes of FT which is why I am speccing them in design consideration today. -
Scott,
I completely agree, it’s all about the sizing. Right now I’m in the throws of a 4 datacenter consolidation\move so really I kinda biased towards the 680 for the memory capacity reasons. In our case that gives us a much higher ROI for VMs\blade using our current (and future) plans with Cisco Nexus and FCoE and with the Flex-10 modules we are deploying. Our storage group doesn’t support iSCSI so we aren’t even architecting anything with that in mind. Excellent points though.
-
Hi Aaron and Scott,
have you or someone else ever considered Sun in this discussion?
If you have a look at their Blade Servers you will probably recognize that there is another player in this league who outperforms IBM and HP in several points. The huge amount of virtualization power can be seen with this Blade module for example:
The X6440 Blade supports up to 4 AMD Opteron Quad-CPUs and up to 256 GB of memory. Additionally you can have up to 10(!)x GigEthernet Ports per Blade and the 10H Chassis can keep ten Blades. Another great thing about Sun Blade is, that it supports 3 different CPU Architectures in just one Chassis. They are Intel, AMD and SPARC.I don’t want go into more detail here, but it would be interesting to know what you and the others think about the Sun way. Cheers.
-
Harry K – There are many other vendors out there (Sun, Fujitsu, Dell, etc) but I don’t have experience with them. I work for a reseller and while we are able to sell just about all companies, our customers have asked for HP and IBM so they are my areas of expertise currently.
I recognize there are other players in the game, I just can’t speak to them with enough confidence to post about them.
-
Wow, that HP chassis sure does look cluttered with management and switching modules. Perhaps that is why 10 fans are needed for adequate system cooling. How much power do the fans require in a fully populated chassis?
-
Brad – Take a look at the previous article I did on power!
-
Aaron / Scott,
I work for a VAR in Australia, and we are starting to see a lot of interest in VDI in the eduation space, for this reason we are building the configs for our Virtualisation with as much I/O as possible, at present the cost between the jump from 6 x 1GB to 1/2 x 10GB for I/O is a little high, but i’m hoping it will drop as more and more 10GB demand is generated.
-
Hi
I have a blade c7000 with 2 x bl460c and a sb600c.
If i configure an iscsi connection on the bl460, will this traffick go through my switchcard in bay 1 & 2 or does this traffick stay internal in the blade? -
@Ewald – I’m sorry to say but I’m not sure. I don’t have experience with the sb600c. I saw specs on it a long time ago (2+ years ago when it was being developed) but I can’t remember anymore. Have you tried the Quick Specs for it? Do a Google for HP sb600c Quick Specs and see if that helps.
Thanks!
-
I know this is an old post but I am working on a C7000 deployment and would like to know if people tend to run a single Dual port hbas in half height blades with out worry or 2 Dual port hbas with 4 FC modules in bays 3,4 and 5,6. I am not worried so much about performance but redundancy. I usually run 2 x Single port hbas in my rack mount servers. In 6 years of working with FC I have only ever seen a port / sfp / LED fail never the hba chip to the best of my knowledge. I also feel that with flex10 replacing a failed mezzanine card or the blade should be quick and relativly painless , no zoning, masking, etc, etc.
Thoughts?
-
Korman, in my experience most people use a single dual-ported HBA rather than giving up the NICs, especially when it comes to VMware deployments. As for how flexible Flex10 can be (and I think you really mean VirtualConnect), I don’t have any direct experience so I can’t comment.
Good luck!
Site Sponsors
20 comments
Comments feed for this article
Trackback link: http://blog.scottlowe.org/2009/02/16/blades-and-virtualization-arent-mutually-exclusive-part-four-hp-traditional-expansion-options/trackback/