HP Joins OpenDaylight Project 37
Mcusanelli (3564469) writes "HP has become the most recent platinum member of OpenDaylight, the open source software-defined networking (SDN) project sponsored by the Linux Foundation. From the article: 'The Linux Foundation, which sponsors OpenDaylight as a collaborative project, is welcoming the addition of HP to the line-up of vendors helping to lead OpenDaylight -- which already includes Brocade, Cisco, Citrix, Ericsson, IBM, Juniper, Microsoft and Red Hat as platinum members -- as a sign of industry convergence around OpenDaylight as the SDN platform of choice. "We are seeing all the major players aligning their SDN strategies around OpenDaylight. HP will be another galvanizing force for the project and industry, bringing the spirit of partnership and collaboration that has made them so successful," Neela Jacques, executive director, OpenDaylight, said in a statement.'"
Re: (Score:2)
"Slashdot are a bunch of sellouts now. There is so much popup bullshit all over the screen, I think i'm done with Slashdot."
slashdot looks better with noscript and the original layout options. of course you deny slashdot the ability to run scripts in your browser that way which imo is good.
Re: (Score:1)
There are scripts on Slashdot? ROFL. Next you are going to tell me there are ads on the internet. (not MY internet!) www.adblockplus.org
Re: (Score:2)
Re: (Score:2)
Um, get a clue you never were their customer, unless you were a paying subscriber.
That is how 'marketing' works. You are the product.
OpenDaylight (Score:2)
Don't really care about HP, but I just perused the OpenDaylight Homepage. It looks very interesting. Does anyone have any more info regarding it preferabley personal experience? Setup/Use Cases etc etc etc.
Re: (Score:2)
Cisco notably doesn't support it except sort of in their own unique 100% proprietary manner, so for 99% of businesses, this is a no-go right from the start.
And HP is different? Looking at the sponsors I just see a big collection of "Embrace and extend" proprietary companies. Cumulus Networks looks a lot more open... http://cumulusnetworks.com/ [cumulusnetworks.com]
Re: (Score:3)
This is likely to make Cisco a no-go once SDN starts taking off. A lot of the demos and projects from other layer 1, 2 and 3 equipment manufacturers use OpenDaylight, and it could give us seamless multi-vendor networks in the future. This is why Cisco doesn't support it, and others like Juniper, Coriant, Ciena do (just naming vendors I deal with, haven't even bothered to check the OpenDaylight list).
Re: (Score:2)
Coriant did an SDN demo at OFC that was pretty slick, building on-demand carrier Ethernet services and controlling both the layer 1 optical and layer 2 Ethernet devices from one GUI. It may be on Youtube or their website.
Re: (Score:2)
Oh, and I'm replying to myself here, Cyan has been doing SDN for years, although not necessarily with OpenDaylight. They've got a centralized control plane for provisioning and operating DWDM, Sonet, Carrier Ethernet, and has integration with other vendor equipment, mostly Overture and Accedian.
Yo Dawg! (Score:1)
I heard you like software, so we put software in your software so you can run software while running your software!
Re: (Score:2)
Re: (Score:1)
It's a fad I tell you.
It will all be pretty much yesterdays news once I finish my 100% hardware based OS. Nobody will be able to infect it in any way, unless he breaks into your home or server room with a soldering iron.
Re: (Score:2)
Interesting (Score:2)
Hmm, that's basically everyone that matters with the huge, glaring omission of VMWare. At this point EMC is going to have to decide between being the 'leader' in the field with their early initiative or being standards compliant and interoperating with everyone else. There was a time where Cisco could go it alone in networking and push their own standards, but I don't think they could today and I certainly don't think VMWare has that kind of clout.
Re: (Score:2)
Hmm, that's basically everyone that matters with the huge, glaring omission of VMWare. At this point EMC is going to have to decide between being the 'leader' in the field with their early initiative or being standards compliant and interoperating with everyone else. There was a time where Cisco could go it alone in networking and push their own standards, but I don't think they could today and I certainly don't think VMWare has that kind of clout.
The Cisco implementation does not work with anyone else. I am betting HP will do the same. Microsoft would if they had a stake in any hardware, but for them it is just a way to keep HyperV relevant.
Re:Interesting (Score:4, Informative)
It's also everyone who has a deep vested interest in holding SDN at bay for as long as possible.
With about 80% of the market, they've got a lot of ability to sneak-in non-standards under the radar. Nobody intelligent would intentionally lock themselves-in to Cisco proprietary non-standards, but it's easy enough for people to buy some new Cisco hardware, and find it has some new value-add capability that companies would like to use. It only just happens to be that it's a crippled and proprietary non-standard way to do things...
The big companies that have embraced SDN, are the same ones who wouldn't use VMWare under any circumstances. The likes of Amazon and Google use open source virtualization software, which they have the resources to modify and add any features they want to add, such as SDN.
Re: (Score:2)
So the same HP that canned its best engineers is a proud sponsor? HP gave up on innovation years ago, so this is nothing more than a marketing ploy and a way to outsource some of its own engineering for a low cost.
No, it is permission to apply a sticker to some switches, and get some press.
Current switches are too limited (Score:2)
The only devices that actually support a useful amount of SDN rules are expensive routers like the Juniper MX series. As long as that is the case, SDN will be limited to doing things that you could do already with a bit more configuration on existing switches. Nice, but not ground-shaking.
SDN will only truly break new ground when someone releases a more flexible switch chip at an affordable price.
Re:Current switches are too limited (Score:4, Informative)
Not sure what you mean by 'useful amount'. Hell, Google has been using SDN for years. Where SDN really shines today is top of the rack datacenter switches like the HP 5900 series. Those cost about as much as a power supply for one of those Juniper MXes.
Re: (Score:2)
More than a few thousand rules. Something that would enable you to actually route particular traffic to particular destinations and back. Something that isn't just a nifty configuration interface to private VLANs.
A few million rules would be great, then you could actually make use of the MPLS tag support. A few hundred thousand would enable multiple rules per VM in a cloud setup. Less than 10000 is just too cramped unless you just use it to make VM migration easy.
Re: (Score:2)
I can see where a large network would need a few million rules, but each individual device shouldn't except for the core of the network, and those are going to need to be on the scale of the MX series anyway to support the bandwidth. Then again, my perspective is from Telecom. Most of my network is relatively few flows, but each at a pretty high volume.
Re: (Score:2)
Most places are OK with 40Gbps at the core and link aggregation, and that is available in dirt cheap Openflow switches. 100Gbps is not quite there yet at the low end, but it cannot be far away now. Alas, long range 40Gbps is still unavailable, but that will probably be fixed within a year.
What makes SDN worth it with a few flows though? Someone must obviously want it, I am just failing to see the use case, apart from easy VM migration (and some cheap switches already have great scripting support).
Re: (Score:2)
I'm interested in it for telecom networks. We've got thousands of miles of network and a lot of techs and 'engineers' that don't really understand the technology. Even for the guys who do understand it through and through, creating or making changes to a circuit/service can take hours today. What I want to see is an application running on an SDN stack that makes it easy to create MEF type circuits, especially across different vendors and technologies. If I need to build an EVPL from A to D, I may have a
Re: (Score:2)
You are pretty much limited to EoMPLS-style service with that kind of configuration. As soon as you go VPLS or layer 3 you need rules either per MAC address or per subnet, and that will eat up your flow table.
Re: (Score:2)
Yeah, I see what you mean. Right now I am pretty focused on EPL and EVPL, but E-TREE and ELAN could be several orders of magnitude more flow entries.