Juniper Networks to Combine with HPE(blogs.juniper.net) |
Juniper Networks to Combine with HPE(blogs.juniper.net) |
Agilent took the heart & soul of HP. Then it split further (Keysight et al)
HP took the printers, which at least used to be a good business.
HPE is the wretched refuse that's left.
Good business as in earning money or as producing customer satisfaction? I don't know about the former, but I do know that HP printers are a horrible user experience these days. The printer is connected via USB, but it cannot be used without an internet connection to their data collection and ink selling cloud nonsense. I will never by anything from that brand again. (I used to be a HP fan since the 1980s.)
Lots of good/reasonable experience with Deskjet 660 and 720C. Although the driver situation for 64-bit windows got real terrible (microsoft had a working driver for server 2003, but insisted on shipping a broken driver for vista/7, so you had to dig up the old one. I think Windows 10 changed the driver model, but I had abanonded the printer for something worse by then. HP support forums could tell you how, but couldn't get it fixed in windows update)
in the late 80's, when Desktop Publishing was a thing.
Cray tends to disagree...
I think I just threw up in my mouth.
This is where I vomited.
HPE would have been purchased by Micro Focus is it wasn't a hardware company. It's really sad to see the decline, HP actually made some nice stuff, 15 - 20 years ago.
Rami (Juniper CEO) is staying and leading.
A major failure of free market economics that is widely ignored by politics.
Does it really matter which OEM you buy your Intel or AMD system from? Desktop computers are a market where the 2 top vendors actually have had all the market until Apple entered the scene with its ARM offerings. I'm excited if others will follow with high-end options (Qualcomm and Nvidia).
I was under the impression Juniper market share had been falling consistently. But it sounds like there is a market where they are still doing well, what is it? Enterprise something?
I think GP was trying to say a few network companies are buying everything up not that Juniper had a majority share itself though. Look at the merger graphs for Cisco, HPE (adding in Juniper's now), and Extreme and a lot of the diversity that used to be "not Cisco" and became alternative growth is going back to being a few companies with a stranglehold on share again.
I would be pretty bummed to change vendor. Junos is probably my favorite networking OS. The CLI is comfortable, BSD is never too far away, and the config structure is gorgeous. IOS (and similar) can really be a pain to read by comparison.
We don't know what HPE is going to do to Juniper specifically, but in general many acquisitions either raise prices or simplify the product line by canceling some products. In this case HPE may cancel some Aruba products and replace them with Mist equivalents.
csco beckons thee
Good lord. I don’t know whether I respect or pity you.
There used to be a saying at Juniper: If there was an issue with MX, it was likely a config issue. If there was an issue with SRX it was likely a bug.
Was incredibly accurate more often than not.
I miss Netscreen.
The Company that Broke Canada - BobbyBroccoli https://www.youtube.com/watch?v=I6xwMIUPHss
"The combination is expected to achieve operating efficiencies and run-rate annual cost synergies of $450 million within 36 months post close."
Not that such layoffs are unexpected.
0: https://newsroom.juniper.net/news/news-details/2024/HPE-to-A...
Aruba wireless...oh man. The support was so terrible.
We had some high availability controllers go out of sync. Couple hours on the phone, no resolution. Had a bunch of back and forth, another couple calls, nothing. It was sort of funny (and sad) watching a different 'tech' try the same scripted steps, over and over again.
Finally I was able to get the issue escalated, and the next guy I talked to seemed to really know his shit. Made some fundamental changes that should have never been that way in the first place, he got one controller back on, and then in the middle the next controller, he fucking bailed off the call and left us in the hands of a new guy.
A new guy who couldn't fix it. He actually broke it in a new away, and then our maintenance window closed. Two more maintenance windows later, nothing fixed, same useless drones trying the same crap, I just gave up.
I ended up fixing the last bad controller myself. My organization was paying almost my salary per year for this 'support' contract. What a joke.
The configuration syntax works as expected, it's flexible but extensible in the right places and mostly has lots of knobs to tune things but without requiring extensive boilerplate to get things to work as expected.
The hardware seems to fit intelligent niches and actually does what it says on the spec sheets.
The system is FreeBSD, while that's a debatable choice in 2024 it's far better than the obscure OSes that others are using for their core systems.
At the end of the day you can always just "start shell user root" and be root on a FreeBSD system. If you're a bit crafty on most Juniper systems you can also run unsigned code if you like. Some Juniper platforms (and not talking about just the huge ones, some of the 1U pizzabox switches) allow you to run a full-blown Linux system alongside the FreeBSD image (because they are actually Linux running FreeBSD on a hypervisor).
JTAC was filled with skilled, intelligent engineers who actually cared and tried to solve your problems. (Not sure today as I haven't had to call them in years). The other TACs I've delt with were focused on call time and ticket handling metrics and would ask obviously pointless questions or repeat information you already provided for a fast close.
I still remember calling JTAC during one outage we had. First L1 engineer after my initial description of the problem basically came back and said something like "that sounds like a really bad outage, please open SSH from 1.2.3.4 and add this public key and I'll login and get the information for L3." Within about 15 minutes I was on the phone with an L3 engineer who correctly diagnosed and proposed a fix for the issue. Amazing support. Not perfect, I have some horror stories too, but, always with people who cared and far less than with C or other vendors TACs.
Very sad day.
Its not really a bad strategy. If you're going to have a large team of network engineers work continuously on a the Juniper tech stack, they'll get use to it - even if they were raised on Cisco IOS. Juniper stuff works just as well as the rest of them.
We bought Juniper gear at the time because nothing Cisco had would work well for us. At least at not any sane price point, and lots of restrictions/gotchas.
Cisco finally got their wind back eventually on their Nexus gear, catching up and run neck-and-neck between Juniper & Arista now.
We continued to buy Cisco for longer than we should have, because "New stuff is surely around the corner", but it's not. Cisco still makes enterprise equipment that can't do IPv6. So we switched to Juniper three or four years ago, and may of our customers are doing or considering the same.
Cisco isn't the dominate play it used to be and is almost never the first choice for new projects anymore.
Not really. Which is why this acquisition price is lower than all their competitors.
> Cisco has stagnated for so long and continue their attempt to push over prices solutions that simply isn't as good or modern as those of their competitors.
Which products, which competitors?
> Cisco isn't the dominate play it used to be and is almost never the first choice for new projects anymore.
Maybe for you, but they are the best in a variety of areas and continue to bring innovation (like Silicon One).
Like any large product company, they have winners and losers, but the existence of losers doesn't assure the absence of winners.
But desktop computer chipsets is the same mess. We can be glad already that there is competition between 2 vendors (1). Some years ago it was probably over 90% Intel.
Servers is not fundamentally better, just with slightly different market shares.
And everything is x86, which is certainly not the optimal solution from several aspects. I wonder whether that would be still alive with some healthy competition between at least a dozen of vendors.
Edit: (1) 2 plus Apple, which is better than only 2, but sti not fully free competition. Few people choose Apple because of their chipset.
Their direction seems to be this framing.. "why would I want to look at a set of graphs/metrics when I can have something tell me exactly what is wrong and make recommendations on how to fix?"
Whether it works in practice or not, who knows. But it is interesting.
Case in point, their 7-layer Burrito [1] offering maps perfectly to the OSI Model [2]. It is fully self-encapsulating (though occasionally leaky), and operates bidirectionally with full duplex support for correctly-configured clients.
[1] https://tacobell.fandom.com/wiki/7-Layer_Burrito [2] https://en.wikipedia.org/wiki/OSI_model
- Digital Alpha: dead
- Digital Storageworks: dead
- Left Hand Networks: dead
- 3par: dead
There's a boot standard, so you can install a range of different Linux-based OS distributions.
Similar in spirit to a whitebox PC, hence the naming.
Google has a zillion links.
Historically, you buy a Cisco switch, you run IOS (or NX-OS).
You buy a Juniper switch, you run JunOS.
A company called Cumulus came around (https://en.wikipedia.org/wiki/Cumulus_Networks) and changed a lot of that (they aren't solely responsible and there are other reasons why but they are very important here).
At the same time, instead of each switch vendor making their own chips, Broadcom started making generic chips which they would sell to anyone (Lookup Broadcom Trident), so, many switch vendors started using those instead of custom silicon.
At the same time, many people started talking about network as code and doing overlay networking things with control planes that didn't fit into the normal models for how these switches worked and were configured (configuration and routing changes which happen in milliseconds over custom protocols instead of ssh to switch and "configure, set route ..." etc etc).
The convergence of this was that the network hardware and network software got decoupled and commoditized. You could buy a Juniper OCX (the OCX line was specifically for this) and install Cisco NX-OS on it if you wanted.
Broadly this was a really good thing for the networking industry IMO.
Also, all of the above generally only applies to the what people refer to as switches (though they can and do route), and not routers (think BGP edge routers) proper (which can and do switch) and are broadly still custom silicon ASIC with much more bespoke and advanced feature sets and can only run the vendor's software.
It does play into their hands perfectly to blame 'the Republicans' or 'the Democrats', and it does indeed make the problem seem more solvable, easy solution right? Vote out the other guy, problem solved, except when we do that, problem not solved, it just changes the guy who's cashing the checks because the problem is much, much bigger than that (and those that cause the problem are much smarter than that).
Also, sometimes, you just want to throw a tool on there to test something.
Today, I'd love to run stuff like tailscale for access to the control plane remotely.
Mostly, however? I did it because they said I couldn't.
It has been years since those features are best served via a general-purpose CPU
Those appliances are nothing but a rebranded servers with some more-or-less interesting software
From there it's a simple manner of looking up who their donors are, and, it should come at no surprise to anyone, that the money that's behind these people comes from entities that directly benefit from their actions.
You are entitled to your own opinion, of course, but not to your own facts. If you don't know the facts - and if you are interested in regulation, especially those facts - then that's your mistake and will lead to other mistakes.
Basically, don't buy an inkjet printer (of any type) unless you really need one for photos or graphics work. And then only buy one if that's a regular thing. Otherwise, outsource to professional print shops in your area. It will be better and cheaper.
For normal office stuff, a laser printer is just easier. The toner lasts for ages, doesn't dry out. And if, like me, you rarely need a printer, it's nice when it still works after just sitting there for half a year of not being used at all.
But a consumer Brother is probably the way to go. Always search the model number and see what people say.
With ~100 pages/year that would have been ~900 pages so far. At 900 pages, that probably would have been fine with the included toner, so toner costs don't count. Assuming you turn it off when not printing so energy costs are negligible. That brings it to ~$0.16/page (with my ~$0.03/page paper) printed.
Printing at the FedEx shop is ~$0.23/page for B&W. The public library nearby has B&W printing for $0.15/page. So with ~100 pages/year its cheaper than a chain office shop but more than the library at this point. A few more years will shift that closer and eventually beat the library price.
I've just given up on the economics of photo printing at home. It seems to me you really need to print a lot of photos to compete with the ease of photo printing at most photo stores in the US. You'd need a really good photo printer to match most of the quality of online stores, so you'd need a pretty good printer with nice ink and paper. In-store pickup at a number of places usually means like an hour turn around or so.
Note though that overall my value for my printer is higher than just the printing. As mentioned, it is also a scanner I've scanned ~1,000 items with it.
They've always worked perfectly on both Linux and Windows and need no maintenance, just remember to turn them on. Most of them are still on their original toner (we don't print that much).
Newer models seem to come with wifi, but I have no idea how well that works.
Laser printers are just better. I can't think of a reason for inkjets anymore that's not niche.
It doesn’t even have WiFi or AirPrint, but brother has an app that can print to it when you’re on the same network with an iPhone, which is surprisingly convenient.
HL-2170W and iPrint&Scan