(From the Archives – September 2009)
News Flash: Cisco is now selling servers!
Okay – perhaps this isn’t news anymore, but the reality is Cisco has been getting a lot of press lately – from their overwhelming presence at VMworld 2009 to their ongoing cat fight with HP. Since I work for a Solutions Provider that sells HP, IBM and now Cisco blade servers, I figured it might be good to “try” and put together a comparison between the Cisco and IBM. Why IBM? Simply because at this time, they are the only blade vendor who offers a Converged Network Adapter (CNA) that will work with the Cisco Nexus 5000 line. At this time Dell and HP do not offer a CNA for their blade server line so IBM is the closest we can come to Cisco’s offering. I don’t plan on spending time educating you on blades, because if you are interested in this topic, you’ve probably already done your homework. My goal with this post is to show the pros (+) and cons (-) that each vendor has with their blade offering – based on my personal, neutral observation
Chassis Variety / Choice: winner in this category is IBM.
IBM currently offers 5 types of blade chassis: BladeCenter S, BladeCenter E, BladeCenter H, BladeCenter T and BladeCenter HT. Each of the IBM blade chassis have unique offerings, such as the BladeCenter S is designed for small or remote offices with local storage capabilities, whereas the BladeCenter HT is designed for Telco environments with options for NEBS compliant features including DC power. At this time, Cisco only offers a single blade chassis offering (the 5808).
IBM BladeCenter H
Cisco UCS 5108
Server Density and Server Offerings: winner in this category is IBM. IBM’s BladeCenter E and BladeCenter H chassis offer up to 14 blade servers with servers using Intel, AMD and Power PC processors. In comparison, Cisco’s 5808 chassis offers up to 8 server slots and currently offers servers with Intel Xeon processors. As an honorable mention Cisco does offer a “full width” blade (Cisco UCS B250 server) that provides up to 384Gb of RAM in a single blade server across 48 memory slots offering up the ability to get to higher memory at a lower price point.
Management / Scalability: winner in this category is Cisco.
This is where Cisco is changing the blade server game. The traditional blade server infrastructure calls for each blade chassis to have its own dedicated management module to gain access to the chassis’ environmentals and to remote control the blade servers. As you grow your blade chassis environment, you begin to manage multiple servers. Beyond the ease of managing , the management software that the Cisco 6100 series offers provides users with the ability to manage server service profiles that consists of things like MAC Addresses, NIC Firmware, BIOS Firmware, WWN Addresses, HBA Firmware (just to name a few.)
Cisco UCS 6100 Series Fabric Interconnect
With Cisco’s UCS 6100 Series Fabric Interconnects, you are able to manage up to 40 blade chassis with a single pair of redundant UCS 6140XP (consisting of 40 ports.)
If you are familiar with the Cisco Nexus 5000 product, then understanding the role of the Cisco UCS 6100 Fabric Interconnect should be easy. The UCS 6100 Series Fabric Interconnect do for the Cisco UCS servers what Nexus does for other servers: unifies the fabric. HOWEVER, it’s important to note the UCS 6100 Series Fabric Interconnect is NOT a Cisco Nexus 5000. The UCS 6100 Series Fabric Interconnect is only compatible with the UCS servers.
Cisco UCS I/O Connectivity Diagram (UCS 5108 Chassis with 2 x 6120 Fabric Interconnects)
If you have other servers, with CNAs, then you’ll need to use the Cisco Nexus 5000.
The diagram on the right shows a single connection from the FEX to the UCS 6120XP, however the FEX has 4 uplinks, so if you want (need) more throughput, you can have it. This design provides each half-wide Cisco B200 server with the ability to have 2
CNA ports with redundant pathways. If you are satisified with using a single FEX connection per chassis, then you have the ability to scale up to 20 x blade chassis with a Cisco UCS 6120 Fabric Interconnect, or 40 chassis with the Cisco UCS 6140 Fabric Interconnect. As hinted in the previous section, the management software for the all connected UCS chassis resides in the redundant Cisco UCS 6100 Series Fabric Interconnects. This design offers a highly scaleable infrastructure that enables you to scale simply by dropping in a chassis and connecting the FEX to the 6100 switch. (Kind of like Lego blocks.)
On the flip side, while this architecture is simple, it’s also limited. There is currently no way to add additional I/O to an individual server. You get 2 x CNA ports per Cisco B200 server or 4 x CNA ports per Cisco B250 server.
As previously mentioned, IBM has a strategy that is VERY similar to the Cisco UCS strategy using the Cisco Nexus 5000 product line with pass-thru modules. IBM’s solution consists of:
-
IBM BladeCenter H Chassis
-
10Gb Pass-Thru Module
-
CNA’s on the blade servers
Even though IBM and Cisco designed the Cisco Nexus 4001i switch that integrates into the IBM BladeCenter H chassis, using a 10Gb pass-thru module “may” be the best option to get true DataCenter Ethernet (or Converged Enhanced Ethernet) from the server to the Nexus switch – especially for users looking for the lowest cost. The performance for the IBM solution should equal the Cisco UCS design, since it’s just passing the signal through, however the connectivity is going to be more with the IBM solution. Passing signals through means NO cable
BladeCenter H Diagram with Nexus 5010 (using 10Gb Passthru Modules)
consolidation – for every server you’re going to need a connection to the Nexus 5000. For a fully populated IBM BladeCenter H chassis, you’ll need 14 connections to the Cisco Nexus 5000. If you are using the Cisco 5010 (20 ports) you’ll eat up all but 6 ports. Add a 2nd IBM BladeCenter chassis and you’re buying more Cisco Nexus switches. Not quite the scaleable design that the Cisco UCS offers.
IBM also offers a 10Gb Ethernet Switch Option from BNT (Blade Networks) that will work with converged switches like the Nexus 5000, but at this time that upgrade is not available. Once it does become available, it would reduce the connectivity requirements down to a single cable, but, adding a switch between the blade chassis and the Nexus switch could bring additional management complications. Let me know your thoughts on this.
IBM’s BladeCenter H (BCH) does offer something that Cisco doesn’t – additional I/O expansion. Since this solution uses two of the high speed bays in the BCH, bays 1, 2, 3 & 4 remain available. Bays 1 & 2 are mapped to the onboard NICs on each server, and bays 3&4 are mapped to the 1st expansion card on each server. This means that 2 additional NICs and 2 additional HBAs (or NICs) could be added in conjunction with the 2 CNAs on each server. Based on this, IBM potentially offers more I/O scalability.
And the Winner Is…
It depends. I love the concept of the Cisco UCS platform. Servers are seen as processors and memory – building blocks that are centrally managed. Easy to scale, easy to size. However, is it for the average datacenter who only needs 5 servers with high I/O? Probably not. I see the Cisco UCS as a great platform for datacenters with more than 14 servers needing high I/O bandwidth (like a virtualization server or database server.) If your datacenter doesn’t need that type of scalability, then perhaps going with IBM’s BladeCenter solution is the choice for you. Going the IBM route gives you flexibility to choose from multiple processor types and gives you the ability to scale into a unified solution in the future. While ideal for scalability, the IBM solution is currently more complex and potentially more expensive than the Cisco UCS solution.
Let me know what you think. I welcome any comments.
maple grove community center
world population clock
isp speed test
breast cancer symptoms
home decorators coupon
good post
what your opinion about HP
are they losing battle to cisco and ibm ???
#HP is not losing the blade server battle in fact they have 52.4% of the marketshare (as of Q4 2009) with #IBM in 2nd place at 35.1%. #Cisco is not even on the chart yet, so the verdict is still out. Thanks for reading!
In your configuration, why not use the Cisco Nexus 4001i switches in BladeCenter H chassis and then manage via IBM Systems Director. The Network manager module in Systems Director presents a single console for all chassis related functions. This will get you reduced cables, reduced # of Cisco 5010s (and thus reduced cost), and simple management? Since Systems Director is a free tool, it would be interesting to see the new winner given this config. Thanks.
Cisco puts the management aggregation point in the interconnect (which has a list price of about $70K) and can manage only 40 chassis at a time. IBM puts that aggregation on a separate management server (IBM Director server) and can scale to a lot more than 40 chassis.
One should separate out management scalability from converged management. Atleast for scalability, it seems like you have left out IBM Systems Director which is typically deployed with IBM configurations.
At the time the post was written (Sept '09) the Nexus 4001i wasn't made for the BladeCenter H. Now that it is available, it would be a considerable option – IF you have the budget and you can't spare to lose ports on your Nexus 5000. Regardless on whether you use the Nexus 4001i or the Pass-thru option, you STILL have to connect to a Nexus 5000 switch. Thanks for reading.
#Cisco vs #IBM – have you tried to use IBM Systems Director for large scale IBM blade servers? I've heard (second hand) that it's not quite ready to mimic what Cisco's UCS management features are.
My experience is second hand as well but I have heard from many sources that IBM Systems Director scales well in multiple customer environments. Also, Director can go across multiple vendors for managing fabric. Can you provide some example on where Cisco's UCS management features are superior to IBM Systems Director?
Good Article, pointing out the inherent architecture differences between the flexible, multi-use case, IBM BladeCenter offering, and the Cisco use-specific design point. I come to the same conclusion that you did, with one minor variation. I would say that there are a number of ways to configure the BladeCenter (as you mentioned) to perform the Cisco specific use of a highly virtualized environment – and it does encompass both hardware (ie: CNA's, pass thru, BNT switch, cisco switch) and the management software (management module with open fabric manager, Systems Director family). However, on the flip of that coin, I don't have the ability to configure a UCS for the more common (at least today) NATIVE OS/application implementations, or mixed technology (ie: mixed speed data and storage fabrics, non converged connectivity, mixed hypervisor environments). So I agree, it really “depends”
Service profiles can be managed across any blade in any chassis. Service profiles contain (but not limited to) MAC, WWPN, UUID, BIOS Firmware, Raid Firmware, Boot Order, etc. A profile can be preconfigured and moved as needed. UCS is equivalent to a 320 Blade chassis. UCSM is GUI (and XML based).
In addition, a UCS system is limited to 320 servers..but you can interconnect UCS Systems with software designed to manage multiple systems (BMC).
IBM Systems Director is basically a spreadsheet with virtual macs, etc managed very manually.
for a first hand experience with UCS read http://www.healthitguy.com
Agree, Cisco is the new guy with no market share. But they also have no legacy systems to support and sell. They get to start with all the latest technology and no legacy.
But the marketshare leader always stands to lose share to newcomers, but IBM and Dell will feel it the most since they have less.
I disagree. UCS is very flexible. It can run any x86 based O/S. The servers are given access to the network…no need to worry about speed (why would you purposefully limit something to 1Gb) and can access any storage resource (NFS, CIFS, iSCSI, FCOE). There is no requirement to use FC storage only. The infrastructure is converged..but what you do above the interconnects is up to you.
How is the many chassis offerings an advantage? Seems like lack of direction and confusion. HP only offers one family (two sizes), Dell only has one. IBM has FIVE??? Weird. If I was a customer I would prefer consistency and compatibility…I doubt you can swap all the components between chassis families which would make support difficult for a data center.
Also – what about cabling. UCS has far less cabling which reduces cost and power requirements in a rack. Let's not forget the upstream switches also required to plug the chassis' into.
Cost – look what goes into a non-UCS chassis: Mgmt modules, switches (Eth & FC) Power Supplies, Fans. This costs a lot of money and the ROI of these chassis' does not come into play until it is almost full! With UCS you have Power supplies, fans and FEX's. This is much lower cost and presents a good ROI for even just a couple of servers. So having a smaller chassis is an advantage for SMALL deployments as well as large.
Server density…you are looking at it too narrowly. Density does not equal # of servers per chassis…but rather # of server per rack. WIth larger chassis' you sometimes run into a limit because of the large power draw from larger chassis'.
I also forgot to add all Storage settings, Network Settings, Server settings are in the UCS Manager. In addition, all user administration and system monitoring is contained in there as well.
IBM's chassis offerings allow for you to use the same switches / management modules and blade servers across all chassis offerings. IBM has a lot more variety in blade server choices, switch module offerings and chassis offerings. For some people this is a good thing. Other people may want a simplified choice. It's up to each person to decide what matters most to them. Thanks for reading!
if you are going to pull in BMC then an apples-to-apples comparison is with Tivoli. IBM Systems Director is free, BMC and Tivoli are not. btw, IBM systems director has won a lot of kudos for its very slick interface 6.0 onwards. Ability to pool resources for management, ability to go across the stack from fabric management to virtualization management are all built in. I guess Cisco and IBM are talking different approaches to management – we'll have to see if the aggregate market likes one over the other.
there is a lot of marketing talk here.
Choice is always a good thing. Unlike Cisco which seems to have decided already that there is only way to architect a data center, IBM allows choice. Why? because it can. Cisco obviously cannot meet the range of market needs given they have no experience in this area.
A lot has been made out about cabling. With the right design, it is possible to get similar cabling with IBM and HP as well.
The biggest issue in your post is cost. You named everything that is cheap but forgot the fabric interconnect which you need even for the blades to talk to one another!
Cisco is also the one with least experience in this space both in terms of technology and delivery. I am not sure why having a legacy system is an issue.
UCS is marketed as built for virtualization but much of that capability comes from VMware and I guess having extra memory which both HP and IBM could do.
i guess we can all make our predictions but how many sales does Cisco UCS even have? who are the customers who actually run production systems on UCS? are the numbers even anywhere close to IBM and HP?
I'm pretty sure Chris wears a Cisco badge. A lot of FUD in your comments about IBM BladeCenter and how choice is confusing. Ask Americans who have choice if living in Dictatorship with no choices would be a preferred option.
Let's face it, if you're looking at Open Systems, why would you purchase a solution that locks you into one vendor? Why not buy IBM iSeries then?
Right now I don't care anymore if my Dell rep is listening or not. They pisses me off with their predatory “promotions” on my personal DFA. In less than a year we will try and go with virtualization. We have a mix of Linux (redhat and MS servers) in our infrastructure and this article is very timely. I might win the support of my management and consider these giants in building my “green data center”
From Dell – “Our records indicate that your promotional plan expired on 06/28/2010 blah blah blah…
Pingback: Captain Harlock
Cisco is #3 now – just passed Dell in the US/Canada.
Still going to ride your horse & buggy to work everyday – cars have no future right? Keep selling yourself as an IBM guy, don’t bother with new technology…stick with big blue. Just like all those guys still supporting iSeries and Mainframes right? Can you say Dying Breed?
How is selling a big, independent chassis with multiple switches in it any advantage? $50k just to house 14 servers (and only 7 servers with the latest X6500/7500 systems.) I can put 8 Servers, in less space, with less power in a UCS chassis at around $10k. So I am not sure why I would want a huge 10RU chassis that costs 5X more with MORE parts in it compared to Cisco with fewer parts and far easier to manage and scale.
IBM Systems Director – what a joke…show me someone that likes it.
Tivoli and HP OpenView – works with UCSM by the way. BMC is just one option.
By the time you read this I could upgrade 320 server BIOS with UCS – could you do that ?
Maybe you fear change, maybe you only know IBM – and that scares you…and it should. Instead of spreading your own FUD, why not look objectively of why there is a better way of doing things. Otherwise, continue being inefficient and selling that inefficiency to your customers. I am sure they appreciate it.
Where are you seeing that #Cisco is #3 in the blade server market? The previous IDC quarterly worldwide server report mentioned nothing about Cisco, yet readers and bloggers alike claim Cisco is #3, so what IS the true answer?
I was fortunate enough to have some communication recently with Jed Scaramella, Research Manager, Enterprise Servers for IDC about this topic. I asked Jed if Cisco is reporting any sort of data in regards to the revenue or units of the Cisco Unified Computing System (UCS) and he replied,
“Cisco has not officially begun to report server sales to IDC (or Gartner I’m told). They have used IDC as “baseline data” in Chambers’s statements claiming the #3 position; we’ve yet to review the details to the claims, so it’s not clear what Cisco is counting and how it relates to IDC’s taxomony of server sales. (i.e., are there including software in UCS sales, which IDC doesn’t). We expect to report Cisco server sales in the next few quarters, after the figures have been vetted and accurately equate to IBM and HP’s business.“
While it does not appear that Cisco is reporting any sales data yet I see on a Cisco blog that on Sept 15, 2010, John Chambers, Cisco CEO, mentioned in a Cisco Financial Analyst Press Conference that a) ”UCS has already taken the #3 market share spot in US/Canada for x86 blade servers” and b) ”Cisco expects UCS to be 50% the market share of the #2 competitor for the worldwide x86 blade server market within the next 2 quarters.”
Now I’m confused. I can’t find any data to support Mr. Chambers’ comments, and for Cisco to not even be on the IDC list of worldwide blade server market vendors, it’s a pretty bold statement to say they are going to be in 2nd place (where IBM is at 24.2% as of Q2 2010.)
Can you show me financial information to back up the #3 claim, because if you can I will gladly write up something on it.
Thanks for your passionate comments, and thanks for reading.
Pingback: KB
Pingback: swellendam bnb
Pingback: Go Communications UK
Pingback: Mustafa El Chami
Pingback: Kevin Houston
Pingback: Wage web » Cisco Blade Server