On 18/10/2010 14:27, Brandon Kim wrote:

Good question Nick, what is a cloud switch? Is this like VSS in cisco where you have a virtual chassis?

The vss is virtual management software for a virtual switch. This box looks like a piece of hardware that you can plug things into, so I'm just wondering what makes this a cloud switch and some other piece of kit not a cloud switch. Nick

Has our industry ever really fundamentally defined what is "cloud computing"????? Even though "MPLS" is sort of a buzzword too, we can define it, how it works, it's protocol and such... But cloud computing?

Subject: RE: Pica8 - Open Source Cloud Switch Date: Mon, 18 Oct 2010 08:26:29 -0600 From: MatlockK@exempla.org To: nick@foobar.org; brandon.kim@brandontek.com CC: nanog@nanog.org

Because 'cloud computing' is the latest buzzword, and their marketing department thought that by attaching that buzzword to it, that would increase sales? :)

Nevermind that clouds contain nothing but vapor.....

Ken Matlock Network Analyst Exempla Healthcare (303) 467-4671 matlockk@exempla.org

-----Original Message----- From: Nick Hilliard [mailto:nick@foobar.org] Sent: Monday, October 18, 2010 8:14 AM To: Brandon Kim Cc: nanog@nanog.org Subject: Re: Pica8 - Open Source Cloud Switch

On 18/10/2010 14:27, Brandon Kim wrote:

...

Good question Nick, what is a cloud switch? Is this like VSS in cisco where you have a virtual chassis?

The vss is virtual management software for a virtual switch. This box looks like a piece of hardware that you can plug things into, so I'm just wondering what makes this a cloud switch and some other piece of kit not a cloud switch.

Nick

-----Original Message----- From: Brandon Kim Sent: Monday, October 18, 2010 7:58 AM

Cc: nanog@nanog.org Subject: RE: Pica8 - Open Source Cloud Switch

Has our industry ever really fundamentally defined what is "cloud computing"?????

Even though "MPLS" is sort of a buzzword too, we can define it, how it works, it's protocol and such...

But cloud computing?

My take on "cloud computing" is simply the provisioning servers or virtual servers (say, VMWare or KVM) on the fly as needed. So you would have a "pool" of servers. When load for one application rises, more servers for that application are taken from the pool and added to the mix as needed. When load drops, that instances are removed from the rotation handling that application and returned to the pool of free (virtual) servers. Providers of network gear have been working on applications that monitor the gear in the application delivery path (e.g. metrics on load balancers) and automatically deploy instances as needed to handle that application. This would be more of interest to providers of "bursty" applications where they might have high load sometimes but a relatively low "base" load. It could also be of interest to people who serve customers in different time zones, such as the US and Europe where the US application can be turned down at night and an application serving Europe loaded up during their business day. It could also be of interest for someone who is expecting a temporary "surge" of activity. It leads, though, to a completely different kind of attack called the "denial of sustainability" attack where a cloud-based provider is hit with a flood of "legitimate" transactions causing the "cloud" management to kick in more servers to handle the additional load. If that cloud is rented, a content provider could be hit with a huge bill.

George:

Nice answer. Do you think cloud services is based on an oversubscription mo= del? Where they hope those who purchase servers don't actually max them out memo= ry/CPU wise?

Do you also believer that cloud services should never have any downtime? To= me=2C cloud services is synonymous with redundancy....

That's an interesting question, and really points more to the fact that "cloud" is rather poorly defined. For example, consider the T-Mobile Sidekick Danger server crash/disaster. This is frequently pointed to as a "failure of the cloud", but in reality, it appears to have been trusting data to a company that wasn't exercising proper care in maintaining its servers. People glommed onto the concept that it was a failure of the "cloud." However, one could argue that quite often, anytime something magically disappears into a part of the Internet we don't have physical control over... I've been toying with defining cloud in a different direction. We have dedicated servers. You get a 10 GHz 24-core CPU with 1TB of RAM. That's pretty clear and familiar to server geeks. We have virtual servers. You get (up to) M GHz and N cores of that same machine. Oversubscription is possible, but not required. In many cases, oversubscription is desirable because that's where the capex and opex savings of less hardware comes in. In both those cases, we get tied up in the specifics of hertz and cores and amount of memory. In the virtual server case, we make some progress towards a model where a VM could be migrated around onto more suitable hardware. This is useful for allowing the proper sizing of a virtual server, for redundancy, upgrades, etc. It seems, though, that ultimately what people seem to be thinking of when they think of the cloud, is the ability to just have stuff "run" without necessarily having to worry so much about the details. In some cases, they're looking for redundancy, or reliability. In many cases, they just want something to be out there without so much effort on their part. They want it to run fast if it gets busy, and don't care if the CPU is oversubscribed ... as long as they can get what they're paying for when they need it. I don't think cloud service purchasers will ultimately be that interested in worrying about whether they "max out" memory/CPU. I think they don't want to have to worry about it too much, though they probably want to be protected from bill shock. That means a model where their server might actually be hosted on a large host with a few hundred other mostly idle VM's, when their VM is idle, and then get migrated onto other hardware if demand spiked. We have technology that can even power on additional host hardware, so there are ways to save on power/cooling during non-peak times. I think you'd find such models are harder to implement if you're too focused on the "evil" of oversubscription. I think what you want to avoid are providers who are unable to maintain sufficient spare capacity to cope with peak demand. ... JG -- Joe Greco - sol.net Network Services - Milwaukee, WI - http://www.sol.net "We call it the 'one bite at the apple' rule. Give me one chance [and] then I won't contact you again." - Direct Marketing Ass'n position on e-mail spam(CNN) With 24 million small businesses in the US alone, that's way too many apples.

We have dedicated servers.  You get a 10 GHz 24-core CPU with 1TB of RAM.  That's pretty clear and familiar to server geeks.

Is that 10 as in Ten?

How does it compare to the OpenFlow design ideas? Djamel On Mon, Oct 18, 2010 at 11:26 AM, Matlock, Kenneth L <MatlockK@exempla.org>wrote:

Because 'cloud computing' is the latest buzzword, and their marketing department thought that by attaching that buzzword to it, that would increase sales? :)

Nevermind that clouds contain nothing but vapor.....

Ken Matlock Network Analyst Exempla Healthcare (303) 467-4671 matlockk@exempla.org

-----Original Message----- From: Nick Hilliard [mailto:nick@foobar.org] Sent: Monday, October 18, 2010 8:14 AM To: Brandon Kim Cc: nanog@nanog.org Subject: Re: Pica8 - Open Source Cloud Switch

On 18/10/2010 14:27, Brandon Kim wrote:

...

Good question Nick, what is a cloud switch? Is this like VSS in cisco where you have a virtual chassis?

The vss is virtual management software for a virtual switch. This box looks like a piece of hardware that you can plug things into, so I'm just wondering what makes this a cloud switch and some other piece of kit not a cloud switch.

Nick

On Mon, Oct 18, 2010 at 10:57 AM, Mark Smith <nanog@85d5b20a518b8f6864949bd940457dc124746ddc.nosense.org> wrote:

On Mon, 18 Oct 2010 13:21:29 +0100 Nick Hilliard <nick@foobar.org> wrote:

...

On 18/10/2010 12:25, Lin Pica8 wrote:

...

We are starting to distribute Pica8 Open Source Cloud Switches :

Sounds interesting.  What chipset does this run on?

Also, what's a cloud switch?  Is this a switch which forwards L2 traffic, or did I miss something?

"Cloud" is the new mainframe i.e. "it's running somewhere else ... "

And the Emperor is naked ... ;-)

Coming back to the subject of cloud impact, from a network perspective, here is an over-one-year-old recording on the subject (Doug Gourlay's comment: "Cloud could break the Internet if not deployed on capable networks"): http://www.infra20.com/post.cfm/fire-infrastructure-2-0-panel-now-viewable-o... ***Stefan Mititelu http://twitter.com/netfortius http://www.linkedin.com/in/netfortius

On Mon, 18 Oct 2010 08:30:48 -0400, Henning Brauer <hb-nanog@bsws.de> wrote:

"Currently, the Pica8 driver is released in binary form"

none of the interesting low-level drivers is open. none. zero.

If it's based on a Broadcom chip, trust me, they are doing the world a favor by not exposing you to the SoC SDK. (It's so horribly un-documented that it took a week to figure out how to build it and another two weeks to actually get it to build something that could be used.)

Yes "cloud computing" is a vastly overused term that is hard to nail down. That is why I try to get people to use the specific technology term they are talking about. Basically in my world "cloud computing" is the vision of having computation and storage just 'out there' without having to worry about it and just paying for a bit more or less as you go... but its not a "technology". Sometimes it helps to flip the term ... for example "desktop computing" is sort of the inverse .. again its not a "technology" really either... just a broad term but we all kind of know what it means by now, like "laptop computing". There seem to be too major views on technologies to make this "cloud computing" happen. Their view on scale/transparency is considerably different. a) bigger layer 2 networks with Vmware type mobility and no IP address changes. Technolgies in this space are much more than just L2 switching, its L2 switching on larger scales with encapsulation, multipathing etc. This is where technologies like IEEE 802.1aq Shortest Path Bridging, IEEE 802.1ah mac-in-mac come to play. These tend to be appropriate for existing enterprise applications (or complete virtual desktops) and simply make existing DC L2 fabrics bigger and availale for virtualization. No application software changes required, its done under them and end hosts can't tell whats happening. b) All new applications/environments that don't care if their IP address changes and deal with it transparently. In this model you can make an application run anywhere, move it around etc without any special infrastructure .. the smarts are between the application and its hosts. Basically dumb infrastructure and smart applications a.k.a over the top stuff. I suppose the hot movement of one of these applications requires co-ordination by the application itself while if its done below as in a) it can be transparent. So depending on where you sit, generically run something anywhere v.s. run specific things anywhere you end up with slightly different underlying technologies... both with the overused moniker 'cloud'. Ok .. flame away .. ;) Peter