Back in 2004, the head of our technical department was lucky to be invited to launch the first Wi-Fi network in Russia. It was launched at the University of Nizhny Novgorod by Cisco and Intel, where before that, in 2000, Intel opened a research and development center with a staff of more than a thousand engineers and even (which is not typical) bought a good building for this. At that time, according to the statement of these two "leaders of production", it was almost the only really working corporate wireless network. Today, such statements about "uniqueness" would probably cause nothing but controversy, but then it was downright a breakthrough.
So it was Wi-Fi of the IEEE 802.11g standard. Of course, the presentation focused on the very fundamental possibility of accessing the network without wires, and here the creators of the technology did not lie, but as for speed and range, there were a lot of innuendo and gaps. Well, actually Wi-Fi G, he is "G", as they called it, that's what they got. To say that they were seriously used in critical areas in organizations would be untrue.
A real step forward was the advent of the 802.11n standard, which became the basis of the starting point for most networks in operation today. History has shown that many people still have N300-type equipment and many of them have enough of it. At least it was enough until the 2.4GHz band turned into a mass radio grave of signals. With the advent of 5 GHz and the 11AC standard, everything has improved somewhat, but apparently not for long. One of the key issues is stability and speed the link is not going anywhere.
Due to the combination of problems and advantages, until recently, we recommended that all our customers connect by wire wherever possible. And this was justified, since 802.11n (comparatively recently called "Wi-Fi 4") did not provide anywhere near the speed and stability that gigabit Ethernet. Of course, with the right installation and choice of cable, which cannot be saved in any case: only good copper and only category 5e or 6. Now we are trying to use only category 6 and +, and it will soon be clear why.
Let's talk about something else. Yesterday we could insist that the client confine himself to a cable connection, but today no longer. The paradigm of the world around us has changed. A quarter, if not half of the devices are gadgets, another quarter are ultrabooks without Ethernet at all (and these are usually all sorts of TOPs and middle peasants that migrate around the office and between offices) and only 30-40 percent of stationary jobs. Therefore, the question “what is this slow Wi-Fi in our office” sounds more and more often. And we are looking for solutions. We try different things.
It was a saying, and the fairy tale is that one of my clients, after replacing the network core equipment and connecting to the “correct” provider via normal optics wanted to improve your wireless network built on Wi-Fi 4 equipment (we will call it by a new name). Over the years, their points have partially failed, so there are many dead zones, and those that remain have already entered a state of complete discrepancy between the capabilities of most of the client devices in use. The word “wanted” in such cases should be understood as the availability of financial opportunities and administrative will - without them, this is just an excuse to talk over a glass of tea. For obvious reasons, I will not disclose the "name" of the client, I will only say that this is a private gymnasium, occupying a four-story building.
An educational institution is a very complex structure, where one is tied to the other, and the issue of access to the local area and the Internet now plays almost a paramount role. Therefore, their IT needs have constantly grown and continue to grow. For example, the administration wants to make online broadcasts of all the holidays that take place in the gymnasium, stream lessons for those who are ill and temporarily studying remotely, conduct group online seminars and teachers' councils with the participation of remote teachers from other branches of the gymnasium. In addition, the gymnasium stores a centralized archive of methodological materials for conducting lessons on servers, which need the fastest access through the intranet web shell and simply through network drives. As a cherry on the cake, it is necessary to give public access for visitors, as parents want to post photos and videos of their children on social networks directly from the assembly hall during the performance of their lovely children.
What we had at the entrance:
~ 15-20% of H~E points gone into oblivion on the N300 standard and a holey coating as a result.
~ 10% of points with "gastritis" - they are, as it were, alive, but they need to be reloaded periodically.
~ very relative "central control"; For the last 2-3 years, the outlets have lived and managed by themselves. Some license was not renewed when the IT administration changed, and it turned out what happened.
That is, 7 years ago, when the building was commissioned, it was a cool network with the latest technology, but something just happened that could not fail to happen: aging of components, overheating due to dust, power surges, hitting the “ball to the point” on defeat, etc.
The number of clients and computerization of the school only grew every year. Cabinets for laptops were placed in the classrooms, and students began to use phones, including for study.
Which is nice:
7 years ago, my colleagues and I also carried out the installation of the wired part of the network, and since the client gave us carte blanche, the cable and connectors were the same Cat6 and a good brand, and a normal core thickness - no hack work. As a result, over 7 years, most of the cable infrastructure has reached more than normal condition.
And all that seems to be needed is to select this wireless part of the network. This is where a lot of controversial issues arise: from the approach to choosing a standard, to brand and budgeting.
Depending on the current moment, the decision to choose a standard can be obvious and non-obvious. Obvious - when the old standard has become commonplace, and the new one is just looming on the horizon. Non-obvious - when a new one is already being implemented, but so far it does not occupy too much of a share.
In this case, the new standard is IEEE 802.11ax, and the old ones - IEEE 802.11ac, renamed, respectively, to Wi-Fi 6 and Wi-Fi 5. Of course, the latest standard network equipment is always more expensive, but the temptation to save money was interrupted by one argument: when we installed Wi-Fi 4, it was also not cheap, but for how many years they worked almost without the cost of modernization, and at the maximum speeds at the time of implementation.
I will not explain here why the 6th wireless communication standard is better than the 5th, many special articles have been written on this subject. Perhaps the only thing that needs to be understood is that we have the same air for all subscribers, that you can’t lay additional air, and each new generation of the wireless communication standard allows you to use the air more efficiently, that is, it provides work for more subscribers at higher speeds.
The next important point is the choice of vendor. The first thing that comes to mind was H~E - it worked and worked well, so we choose something from H~E / A~a.
We make a request to A~ac AC and from AX. This will be A~a N~s AP-5~5
We get: Ar~ AP-5~5 - with AH - 63 thousand rubles (November 2019) and A~a N~s AP-3~~ with AC - 52 thousand rubles. (November 2019). We need such points per object (4 floors of 10 x 15 pieces = at least 40-50). Total: 2,6 million rubles if you take 11AC in RRP prices. About 11ax, it remains to say only AX how expensive it is and put it off for later. And we have not yet reached the cost of the controller and licenses!
What happened in 7 years? And the rate has gone up! Then, in the 13th year, brand outlets also cost under 600-800 dollars, but the rate was different. Although the gymnasium is private, it receives income in rubles. And then there was a cognitive dissonance and rethinking at the stage of discussion with the customer.
Everyone knows the concept of overpayment for a brand. And in this case, this option is clearly here. For a client, the choice of a brand means one thing: if you don’t understand, buy from the most promoted ones, then you won’t make a mistake if you can pay of course. For us, selling an expensive product is also great - we will earn more. There remains the risk that the client will “skip” to someone who dares to offer something cheaper, because we and the client are in 2020, and not in 2013: the crisis is behind us, a new one is on the threshold and we need to think with our heads.
So what do we do? Persuading the client back to forget about AH? And if already AH as you want?
So we are looking for options!
Fortunately, the IT market is dynamic: something constantly dies and something new appears on it. Sometimes, newcomers, in an effort to captivate the public, provide the same or similar characteristics as the "A" brands, but for less money. Of course, there is the risk of lottery, roulette and even Russian roulette with a loss shot. But it can be relatively minimized if you carefully approach the filtering at the level of thorough tests before purchase.
What is the probability of finding a golden ring in a pile of last year's foliage? The answer is 50/50% - either you find it or not - most likely not. But it happens to be found.
We, as integrators, are invited to all conferences. In my opinion, about everything: from telephony and intercom to access control and Wi-Fi. Sometimes we go. In addition to marketing, in 1 out of 100 cases there is also a healthy grain.
Just last summer, a certain Taiwanese EnGenius participated in a similar team conference like “salad from different vendors”. Who is not clear. All that remains in the memory after a year is that the brand is similar to the name of the manufacturer of mice and declared Wi-Fi 6 ready to work, aka AX. I remembered miraculously, just looking at the Genius mouse.
Got to their website. I dug up a presentation from the distribution of that conference in the mail. When studying the slides, it turned out that EnGenius is allegedly a contract manufacturer of network devices (in particular, access points and controllers) for brands such as Cisco, Dell, Extreme, Fortinet, Zyxel and someone else. If you believe the Taiwanese, then at the same factories with the same technologies they also make wireless under their own brand.
In general, it turned out that EnGenius has long had Wi-Fi6, as they do for the "senior". Moreover, they were almost the first in the world to manufacture Wi-Fi 6 (IEEE 802.11ax) network devices.
A year ago, this was just curious information that was quickly forgotten, but now, when the issue of modernizing Wi-Fi in the gymnasium has risen sharply, it has “fired”.
Question number 2. How much and where to get samples.
The first thing to compare economic efficiency. A cursory appraisal of the retail gave a stunning effect. Point with AH from Engenius costs on average as half of the branded class "A". So the problem is inside! Or, again, the overpayment factor for the brand?
We need samples. Without in-depth testing, a product with such characteristics and such a price should be considered, as it were, “dumb”. We call different companies in Moscow and St. Petersburg - there is no product, but who accidentally is, he does not give tests. There is no more talk about AH points.
But we are stubborn! We write in Taiwan. For some reason they answer from Holland. It turns out that there are Engeniusists right in the Russian Federation. After correspondence in school English about empty clarifying conditions, we get the contact of people in Russia. It turns out that there is a test fund. The product exists and you can take it to touch.
After describing the problem with an emphasis on choosing only points with AX and signing gar. letters, a week and a half later (from Samara!) received a set of 4 different points, including the AX-naya and the PoE switch, which also turned out to be the network controller.
Taking into account all factors (price limit, required density and gymnasium's wishes), EnGenius EWS377AP access points were chosen for tests and potential future installation.
What they looked like: the speed is declared up to 2400 Mbps at a frequency of 5 GHz + 1148 Mbps at 2,4 GHz. That is, this is a plane, according to the numbers.
The kit came with an 8-port gigabit switch controller with PoE +.
Of course, it was suitable for tests, but it is obvious that the traffic potentially generated by the AX point cannot be transmitted through the gigabit Ethernet port. Actually, the point itself is immediately equipped with a multi-gigabit 2,5 Gb / s interface. If anyone remembers, this was adopted back in 2016 and is just starting to take effect.
In principle, this feature of the points fit exactly into the theme for the customer, since after the upgrade of the network core at the gymnasium, most of the ports are just multi-gigabit over copper + part of 10G SFP +.
Everything is fine, but this raises the question of choosing switches. In the case of EnGenius, if you make a homogeneous network, only 8-port switches with 2.5G with PoE + are available. Initially, we planned to lay a 48-port high-density, or on the edge of 2 x 24 ports with uplink via SFP for growth in the number of PoE + ports. But so far EnGenius has all gigabit ones like the eight-port one that came in.
The good news is we can brag about our favorite cable theme. The presence in the project of cables of the 6th category, laid "for growth" and capable of passing these 2,5 Gbit / s, greatly speeds up, reduces the cost and facilitates the task.
As we can see, at the time of laying the cable system, there was no active equipment with such speeds, and this once again confirms that there is definitely no need to save on cables.
As a result, the picture is as follows: we are testing the system on their 8-port switch controller, but in the future we will probably take as floors. Details of the required number of ports will be shown to us by radio planning.
Step 1.
We collect the stand from the sent points and the Switch-Controller.
We go to the web interface.
The main page of the switch, aka the controller.
Distribute the points into groups.
Great! The entire network, including other people's devices at a glance! Convenient and affordable.
Step 2.
We are looking for a radio planning tool in the controller and do not find it.
It turns out that EnGenius has radio planning, but it is moved to the cloud and is called ezWiFiPlanner. We call tech support comrades from Engenius. We are registered in their system and given access.
So what do we see here.
The cloud-based Wi-Fi coverage planning system is extremely powerful. I would say that Ekahau claims to be samples of similar products, but with only one pleasant exception - this ezWiFiPlanner is free. From the word at all. The downside, of course, is that she knows nothing but her EnGenius points.
A simple sketch of a radio plan can be done in a matter of minutes, which is what is done in the video. Then it remains to circle the walls and windows, indicate which walls are load-bearing, and where these are plasterboard ceilings. We clarify with the customer that we are fixing the points on the ceilings and as close as possible to the previous places, moving back and forth and the points occupy the final places.
In general, I can confirm from my own experience that working with the EnGenius scheduler is quite easy and convenient, the libraries have everything you need. It is easy to change the network parameters and we can immediately see the result. I note that you can save your projects in the cloud, and then export them and use them as templates for other objects. This is a plus, as I have seen from my own experience that many built-in software in controllers do not allow you to save the radio plan and even paid systems that do not allow you to export the project to a simple PDF. What did they pay for then?
Well, here we get such a coverage scheme for our object
This is the layout of the first floor for a frequency of 5GHz, the rest of the floors have an almost identical layout.
Actually, that's the whole solution.
As for the choice of places for installing access points, in our case it would be possible not to be smart and put new access points in the same places where the previous ones were installed, the Wi-Fi 4 standard, well, maybe cover the assembly hall more tightly. In fact, we did just that, trying to minimize the work of re-laying cable routes to points. Nevertheless, in view of the resulting real picture of the new radio plan and the list of wishes / improvements from the client obtained by the experience of 7 years of operation of the previous network, part of the cable ends still had to be re-laid in other places, and part of the sections still had to be re-laid along the trays. But in general, this can be considered a minimal upgrade.
When planning, I preferred, as the preferenceists say, to re-mortgage - the number of client devices and the amount of traffic will only grow, and I would like this network to also stand idle longer without the need for modernization.
Step 3. Testing and comparing.
It is time to understand what the same AH gives us. Moreover, in addition to the AX points, we also have Wave2 + Wave1 with different configurations of antenna circuits. So we can afford to compare results. For tests, we take the Samsung C10 with the declared support for AXa (802.11 a / b / g / n / ac / ax 2.4G + 5 GHz, HE80, MIMO, 1024-QAM)
Measured on EWS360AP and EWS377AP.
The tests were carried out at a distance of 2-3 meters from the point, i.e. the typical distance from the point to the student in the class. On the latest Galaxy of one of our techies, we once managed to get almost 640Mb / s over the air. Which is basically impressive.
The results are very interesting ~320MB/s on EWS360AP(AC) vs. ~480MB/s on EWS377AP(AX) over LAN. The increase is as much as almost 50%. Naturally, in real conditions, the speeds will be less, but the difference is absolutely obvious.
Surprise where you didn't expect!
We will assume that our tests are very similar to positive ones. It remains to resolve the issue of managing the entire network, as part of a combat project. The EnGenius EWS377AP access points planned for use certainly have a built-in web interface for configuration, but it makes sense to use it only for single use, outside of a group. We have another task - conducting an entire matrix of points.
On the scale of the gymnasium, it is necessary to obtain seamless roaming according to IEEE 802.11k / r / v standards, and a guest network separated from the main one, and possibly more than one. Basically allow you to make as many as 16 SSIDs, with their own group policies (for administration, accounting, teachers, students) - but this is all possible only with centralized management.
While working with the Engenius switch controller, I kind of got used to the idea that the PoE switch and the controller are one and the same person and you won’t need to pay extra for anything. However, when moving on to compiling a specific specification, we found that the new 2.5GbE PoE + switches from EnGenius do not have an integrated controller, as they are hybrid - local-cloud. It is assumed that in the future we can move from local to cloud controllers. Maybe this is a global trend, but so far such an option would only cause panic in the client, so the TP asked what other options are.
In response, 2 options were offered: installation of a free product to a computer or buying a hardware mini-controller with functionality and web interface identical to ezMaster.
Let's summarize the problem of choosing a platform in a table
SkyKey - mini controller
ezMaster - software on the server
Maximum number of points in an array
100
1000+
Management
Via EnGenius Cloud or web interface locally
hardware requirements
The controller box
Media required: PC or server and virtual environment
System startup speed
Almost instantly - stuck and go to work
You need to figure out how to install, configure and then as always ...
Certain fluctuations took place, but here, too, they decided to follow the path of simplicity in the combat project. Plugged it in and took off - Plug-and-fly!
According to the practice of life, a specialized network device with a web interface, which can be accessed from any client (including from a smartphone), always captivates with potentially greater stability than the installed software, especially for a very specific deployment - through VMware. I have nothing against VMWare, the virtual machine has its advantages, but it must be done or implemented on the gymnasium server that virtualizes other tasks. This time. And in principle, we save money for the client quite well.
Limiting the mini-controller to 100 access points for the gymnasium is not critical - in our wildest fantasies we will not even come close to the limit, and radio planning gives us less than half the load.
The magnetic fastening of this contraption put an end to the discussion. Clap! - stuck. Everyone laughed and took it.
Schematic diagram and core of the network.
The general connection scheme turned out like this. The total number of points decreased from 40 to 32x.
Since there are 4 ports on the "main" switch, and we would need 5, it was decided to connect the third floor through the second one (half of the second floor is occupied by assembly and gym halls and there are much fewer clients there).
And Juniper EX2300-24T was chosen as the core of the system. The choice was between him, SG500X-24P and AT-GS924MPX-50. But with the closest possible characteristics, the device from Juniper greatly wins in price and fits into the budget.
Summary of experience gained.
It's too early to draw conclusions. Conclusions can be made only when the network is finally put into operation and will work for at least six months.
While impressions can be divided into 3 components.
Positive:
- The price for AH is more than adequate. In fact, the choice of this particular vendor made it possible not to abandon the idea of taking Wi-Fi6 in principle. If you look at the rest of those who already have AX, it's expensive and there are a lot of confusing licenses. For example, I have great respect for the company A~d T~sis, but taking money for seamless roaming is a game and horror in our time.
- I was very pleased with the Wi-Fi glider in the cloud. Made more than on the level and for free.
- The controller interface is quite correctly made, where the entire network is transparent and everything can be controlled through and through from one screen.
- the appearance of the points is neutral, they dissolve in the interior, the brand name is almost invisible
- despite the fact that we directly risked choosing something unknown, the network on Engenius works. And it works with a bang. The signal is stable, does not jump, dots do not fall off. Time will tell how they will behave at a mass event in the assembly hall, but the entire segment of the office, which was launched first of all, lives very steadily.
- roaming. He is. I will not prove that it really works, as we did with another product at the dawn of this phenomenon - it's just that many have it in our time and any normal manufacturer should have it that way
- Native support for Mesh networks and its configuration did not cause any problems
+ band-stirring. Yes it works. Normally flips between ranges.
Negative:
In my opinion, a stupid and not very reliable mount to the ceiling. For many cheaper even products, the mounting base is made of metal and it is fixed tightly and is more convenient to mount. I don’t say anything, it holds, but here we have an “active contingent” with mass runs along the corridors and throwing objects into the distance, so we expect potential problems.
The cable entry window at the 377th point is, to put it mildly, not very reasonable. The cable must be introduced from the ceiling to a break, and if you supply power from 12V with a separate pair and a clamping chip, then it barely fits into this opening. The situation is aggravated by the rather “non-blunt” edge of the metal back, which can crush the cable.
Neutral Strange:
It looks somehow strange that in the older version of the gigabit switches there was a built-in local controller, but in the new ones it is not.
Finally.
The choice of whether it is necessary to run right today to buy AX points remains on the customer's side. It is clear that this is a trend. If there is a trend, then you should not blow against the wind, but bet if asked.
Purely by technical advantages, you can judge and draw conclusions yourself based on the test results. The huge question is what will connect to this Wi-Fi6 and how to measure the speed. On old equipment - zero reason. But new - the increase is obvious if the network coming to the point is also adequate.
It remains to answer the question, what is EnGenius? The general impression is more “Yes” than “No”. What bribed me was that the network rose all at once and without tambourines, and everything flew off. But we will be able to judge in general in a year. For now, let's put an ellipsis, but nothing really bad can be said.
Situation for now.
During the period from the beginning to the middle of March, we managed to put the pilot segment into operation. Now, for obvious reasons, we cannot continue the deployment of the segment, but the test results obtained are more than encouraging.
Source: habr.com