All I can think of
It was about damn time!
Networking giant Cisco lost around $10bn in market value last week as investors sold out in response to weak earnings and poor growth projections. Worse yet, some of its biggest customers – from Amazon and Google to social network giant Facebook – are implementing technologies that will allow them to reduce their dependence on …
It was about damn time!
It has been a while since I was involved in such a purchase, but it is maddening. We had no choice but to go with vendors like Cisco due to performance issues, but I have *never* liked it. It is not just *more* it is a crazy amount more. I am sure it is better now, but at one point I did a little course on Cisco stuff and programming them was like throwing switches on ancient iron from the 1960s.
That type of infrastructure should have had a price/performance ratio that *at a bare minimum* tracked Moore's law. Bandwidth in all its manifest forms will remain the primary bottle-neck for the foreseeable future. Cisco and friends should have been well on top of this, but instead of helping to build out the network to the enormous capacities we need they have concentrated on margins and net-profits.
I am not sure of the case with Cisco, but I know that JDS (RIP?) outsourced their manufacturing to China as soon as they were able, destroying high-tech jobs in Canada and eventually the company too.
I expect Cisco to get spanked. In its death throws it will likely start hauling people into court, but I doubt, even with our patent system, that they will be able to stop others from building out the network.
Next stop? Probably an attempt to shoe-horn in Cisco patented equipment for 'National Security'.
"We had no choice but to go with vendors like Cisco due to performance issues"
just about everyone in the market outperforms Cisco by a wide stretch these days, even at significantly lower price. It must have been quite a while ago that you found that problem.
Wouldn't immediately dismiss Cisco patents. Could see them being awkward around licensing and they do spend a fortune on lawyers.
Don't get me wrong if these efforts succeed then it will cost them a fortune in revenue. But I would bet they still make more from licensing than the OEM's.
We are getting closer to world dominance by a handful of Corporations every day... watch out for the Aliens
Been saying it for years - Cisco is going to lose the switching market, and probably the routing market after that.
I see them being relevant in corporate IT (Voice, user access) and servers (I'm guessing that their switching market share will collapse into the embedded switches they're building into UCS chassis).
I'm not 100% sure that home grown devices are mature enough right now for the regular DC. However from an OS perspective, Cumulus Networks should change this in the coming years, and hopefully a larger install base of Trident I / II based boxes will also level the playing field.
Interesting times ahead, and not even one mention of SDN (d'oh!)
I never understood why Juniper or Brocade weren't more aggressive in their pricing strategies. We trialled some of Juniper's kit a few years ago, and found them to be cisco in a blue coat. I was prepared for the higher fail rates and clunkier hardware that i got, but was really surprised when i saw the numbers we were paying. We ended up going back to Cisco.
There's a real demand for some proper competiton in this arena, and yet the other players just don't seem to step up.
"I never understood why Juniper or Brocade weren't more aggressive in their pricing strategies."
They are. Both of them came in with comparable kit to Cisco at 1/4 the price in a recent prepurchase round.
meh, we quoted up on a few hundred EX4200x, a hundred EX4500's and 25 8500's. They were within 15-20% on cisco and i was disappointed. hardware was cheaper, service contracts weren't much at all
What surprised me is that none of the commenters so far have mentioned Huawei (or even HP).
We're going the Huawei way.....better, faster, cheaper.
And support second to none.
Title says it all
I use Huawei and ZTE. Why not? America is no more a friend of Canada than China; I've zero reason to trust either. Given this - and that they both have performance good enough for anything I could ask of them - I might as well buy the cheaper kit.
"Buy American" is bullshit. Both nations seek to be my nation's Master...but we only share a border with one of the bastards. I'll bung my $ at the nation that actually has to work to take us over or cow our politicians into submission.
One of cisco's support headaches for many years was that practically every major customer had a tailored version of IOS to suit their specific requirements. An open architecture would at least pass that problem back to the customer and give cisco an opportunity to retire a lot of the feature bloat that has accrued as a result (which, I suspect, the engineers - if not the salesmen - have been aching to do for years).
(disclaimer - I've never been a Cisco customer, and never will be)
Those big cloud players have enough purchasing power to likely get Cisco gear at the same cut-rate price as the competition offers to other players. I remember one big Cisco deal many years ago where the customer got 90% off Cisco list pricing just to keep the competition out (Cisco had otherwise lost the deal on all fronts whether it was performance, cost, etc).
I recall again many years ago(5+) that Amazon had been building/deploying their own custom "TOR" switches with I believe broadcom chipsets. probably goes back further than that. Though big core switches were probably from one of the big players, same goes for probably google and others too. I don't track networking too closely but have not heard of any large scale "white box" switches - e.g. that do hundreds of 10Gbit ports at line rate speeds for aggregation/core workloads.
Look at the average cost of a server today vs 10 years ago and likely you'll find it quite a bit more expensive. ASP is up across the board on everything I think, though what you get for that price is also significantly more than it used to be.
I remember when a 4-8 port 10Gbps blade for a core switch would run easily $40-50k+ (and to even consider 10Gbps at the time you had to have a chassis switch) -- now we get 1U buggers with 1Tbps of fabric and 56 ports for under $20k, or 48 ports with some minor restrictions for well under $10k - in almost all cases that little 1U 10GbE switch has significantly more fabric capacity than the 10-20U+ core switches of the earlier era), and obviously a small fraction of the power draw.
What you used to pay a quarter million for on a core switch to do ~100-500Gbps of switching fabric you can easily eclipse 10Tbps today.
So, yeah. A bunch of whiny little (or I guess big) bitches.
You'd be surprised. My company buys a lot of Cisco gear (approx £2mil of switching/routing/VOIP gear here) and we get a very marginal cost reduction; Juniper are at LEAST half the price. We recently renewed our TAC account and support with Cisco and decided not to replace our Datacentres with Juniper, even though we're paying about £1million more with Cisco over the lifetime of the kit.
Couple that with the fact that we are continually getting frustrated with being overcharged for things like spoke-site routers because we have to buy the really expensive top-end IOS featureset to get one protocol to work (the router does VPN but no tunnel encryption unless we pay another £300! Pointless!).
Then comes the frankly insane support costs. And the fact we cant even get a free copy of Packet Tracer to build virtual labs and test new designs for our enterprise!
Cisco need an open source like this to give them the much needed kick in the valuable to change their thinking. Either that or they eradicate them entirely, which I think would be the preferred option tbh.
You're somewhat incorrect in your facts..
> I recall again many years ago(5+) that Amazon had been building/deploying their own custom "TOR" switches with I believe broadcom chipsets. probably goes back further than that.
Not TOR, rather leaf/spine. More like 3 years ago. Correct in Broadcom (trident based).
> Look at the average cost of a server today vs 10 years ago and likely you'll find it quite a bit more expensive.
Absolutely correct, and absolutely incorrect at the same time. Cost is probably linear for the high end stuff, however performance/$ is now WAY up. ie: you dont pay top dollar for 10 year old servers.... Hence the comment about price following moores law.
However in the networking world, we only go by jumps of 4x or 10x. This happens at very, VERY slow rates, and the associated price/port is insane, and very slow to drop. So if you model it at price/performance, networking is lagging and expensive.
Since Cisco develop their own ASICs (well, most of the time), they cost more per port than vendors who use merchant silicon (ie: Arista, Juniper). They benefit from economies of scale, and not being burdened with the cost of developing the hardware. They just get the specs from the manufacturer, and off they go, predictable hardware development costs.
> I remember one big Cisco deal many years ago where the customer got 90% off Cisco list pricing just to keep the competition out (Cisco had otherwise lost the deal on all fronts whether it was performance, cost, etc).
Not uncommon, from what I see, these days, it's the only real way Cisco can compete.
> What you used to pay a quarter million for on a core switch to do ~100-500Gbps of switching fabric you can easily eclipse 10Tbps today.
"Back in the day" is not today. Nor is it a valid strategy for planning a network that is experiencing exponential growth.
I´m living in the world of public tenderismus, i.e. the cheapest win (well, at least they say so). And the winner for us was: HP. Why did they win? No, they don´t have the best hardware, no they don´t have the best software (way no), but they win because of pricing.
Of course, when I read about the next super Nexus / MX / what ever switch with Terabit of throughput, I´m accided. But in my world they will never find their way into my "company".
The guy doing the tender knows "OSPFv2" and hey he heard about "OSPFv3" and IPv6. Btw, that was everything - well wait I forgot to mention "dual stack". when it comes to the tendering document. God, we could have something like Netgear, but thank god they didn´t take part in the last bidding. 9 million € within 4 years for a large group of different public institutions. Seems not to be that interesting for them. Also, Juniper didn´t take part from what I heard nor Cisco. But it was actually hard for me to get those information. Everything top secret, it seems.
Well for me in my smaller institution, budgets have been again cut, now I even go cheaper and I go refurbished hardware plus some fanless switches from Zyxel. Don´t laugh they work, too.
Coming to the point, nowadays everything is about pricing. May it be good for the user or bad.
I think it's fair to assume with exponential growth comes exponential costs, there's only so far down you can drive stuff. You can bend the cost curve a bit, but if your already purchasing probably $10s-100s of millions worth of networking gear per year you already have a lot of the benefits of such scale already.
Now maybe Cisco is able to gouge these cloud players as much as they can the little guy, I don't know, in which case they probably only have themselves to blame as there have been good alternatives to Cisco going back 15 years now.
Even using IOS makes me want to gouge my eyes out with an ice pik, it's so needlessly complicated.
I'm not completely sold on companies like goggle, facehack and others building their own networking kits and selling them to the rest of the world. Too many backdoors can be built in (and are, for that matter) and enabled for themselves to glean information from unsuspecting people when they build it. I've used Cisco for years and, while I think they charge too much, I'm not entirely ready to let go of the proprietary security they can (and do) provide.
We dumped Cisco at all of our locations years ago. The cost was insane (a $150 for a freaking NIC?!), the never-ending expensive certifications, and the arbitrary limits built into the Cisco system all were a huge drag. Worse, if you wanted to do something more you had to upgrade not just software, but also hardware. I built our first Linux router and never looked back. We now have up-times of a year or more except for the random long-term power outage or the need for an administrative reboot. We have complete control of the entire system. We can add or remove features as we see fit. We can put it on hardware that fits our requirements. We can train our own employees. We can build-in API's for monitoring and command. I control my network now, not my vendor.
If someone in my company bought Cisco, I'd fire them.
That only applies at small scale
There is no PC on the market today that can cope with a real workload, e.g. 3xOC192 and multiple 10GBE links. The PCI Express slots just aren't designed for that.
From various people who work inside large enterprises, Cisco appears to have lost the plot quite a while ago. From personal experience, the 65xx chassis was massively underprovisioned in terms of backplane speed for the port density it provided (and this was back in 2002 when I worked at a large ISP. They haven't revised the backplane since AFAIK).
Cisco's answer? Push behaviour modules. Want your firewall in a blade? How about your IDS and IPS? Oh, and your e-mail hygiene product too!
They tried to cover up the lack of backplane bandwidth by pushing stuff that really should not be in a chassis slot in the first place. With the benefit for Cisco that they get to sell more switch chassis also!
The "cloud" companies moved off Cisco a while back, especially Google. If you take a minute to think about the way Google FS works, you'd realise why.
Not sure I agree onk and Amazon are at least as far in the NSA/GCHQ pocket as Cisco is.
So why does the CIA award $600 million to Amazon? Well, them being one of 20-30 agencies in the world exempt from 5 eyes spying has a lot to do with it. If they can reach an agreement with Amazon to keep instances in datacenters located in 5 eyes countries, then they are probably pretty secure.
Everyone else? Much less so.
Really, they gouge you on everything they can, to the extent of locking out rival SFP units "because they can":
If they had a supported list that included any rival's products then you might have believed the quality assurance bullsh*t...
CISCO saw this coming a few years back and thought about diversifying, so what did they do? Jump into the cut-throat server market. Out of the fryingpan into the nastiest cooking fire in the industry!
I think this is more whining than accurate.
I do consulting for a very large organization, and our current discount is ~60% off MSRP for all customers.
Cisco may be hamfisted in not lowering the price for smaller/SOHO customers though. However, the Facebooks and other large companies are certainly getting 70-80% off of list, which is an insanely good price.
Not that Cisco doesn't need to be taken down a peg for their ridiculous<sp> MSRP and backroom discounting business process, which should have died at least a decade ago.
But these guys are like Apple, they are used to beating their vendors down into the dirt on price.
Really, 70-80% off MSRP is not a good deal, its quite possibly barely above Cisco's actual cost + overhead...
Good luck trying to figure out who too call when your enterprise has a WAN or other segment go down. Do you call the equipment manf. or some other entity?
Since when has it been the case that the tech industry has not made its money from presenting the marketplace with a black box which is just another phrase for lock in.
"They have closely guarded the inner features of their technology, and every release with it brings a wealth of new features for their platforms that requires a legion of people armed with certifications and arcane experience to manage."
As one of the army with the certifications and arcane experience, I don't see this being improved by what is happening here. Even though arcane at least networking skills are portable between companies becuase people use the standard kit and features. If companies start creating their own networking kit with their own unique features, it is going to make the sector even more arcane and it's going to be more difficult to get staff to support it without a lot of internal training first. I do not see that as a positive.
Otherwise competition for companies that have a bit of a monopoly is good.
These 3 companies are Arista customers. Obviously they have seen the light.