PhoneBoy’s Security Theater

If you're seeing this, the DNS hasn't updated yet.

I was listening to Episode 395 of the Security Weekly podcast when I heard one of the guys say that Information Security is often The Department of No. As in, no, you can't use that device. No, you can't use this new technology. Why? Because we can't secure it or control it. Because the risk of data loss is too great. Because NSA. Or whatever boogeyman The Department of No can come up with.

There are points on both sides of this debate. Anytime a new device connects to a network, regardless of what it is, there is always additional risk. If that device is controlled by the company, the risk is significantly less. If they don't control it, and there are inadequate security measures built into the network they connect to, it's another potential member of a botnet or worse. 

Despite all these risks, people want to use whatever device they want wherever they want and still get to and work with corporate data. Most users have zero clue about the potential information security implications of this, and even those that do--often those folks who sign paychecks--don't care because they want what they want, implications be damned.

So what's a poor information security professional to do? How can they become The Department of Yes while minimizing the associated risks these new technologies present?

First, we should ask the question: what is it we're trying to protect, anyway? Certainly an unmanaged devices presents a risk and should be connected to an untrusted network, such as a guest WLAN. You have your network properly segmented to allow this, right?

Next, you should ask the question: what is it that device needs access to? Certainly not the entire internal network, but some subset of it. Common items include Email, some intranet sites, their home fileshare, and maybe Salesforce. 

Of course, the minute you provide access to that data to an unmanaged device, there's a risk for that data to be lost somehow, which should raise the hackles of any information security professional. The common reactions to this phenomenon are to not allow access to any data, only allow access to very limited, harmless data, or to allow it only on condition of managing the device (i.e. with a Mobile Device Management solution). Users, on whole, like none of these options.

Another option is available, and that's the concept of a secure container. This allows an unmanaged device to access sensitive data within a container that is secured and controlled by the business. The user can do what they need to do, the data only stays inside that container, that data is stored securely on the device, and the business can revoke access to that data at any time without managing the entire device. 

The challenge with any container-type solution, of course, is doing this without compromising the end user experience. The users still need to be able to access and potentially manipulate this data in an intuitive manner. If adding a container-type solution degrades the end user experience, end users won't accept it and will demand a less secure solution. 

There are a number of different solutions that operate on this principle. The ones I am most familiar with are the ones Check Point sells. Given I work there, that should be no shock to any of you. They include:

• Mobile Access Blade, which provides a way for unmanaged desktop/laptop systems to access corporate and manipulate documents within a Secure Workspace. This workspace is encrypted and disappears from the local system when the end user disconnects from the gateway.
• Capsule, which enables similar access on mobile devices, as well as provides a clean pipe to mobile and desktop systems that are off your corporate premises by routing traffic to Check Point's cloud where traffic is inspected using Check Point's Software Blades using the same policy you've already defined for your enterprise environment.

Regardless of whose solutions you choose, they allow you to be The Department of Yes. Yes, you can access that data with your chosen device without the enterprise losing control of the data. Everyone wins.

While I don't typically read End User License Agreements, I was pointed to the following phrase that exists in the End User License Agreement for Palo Alto Networks equipment:

1.3 License Restrictions: End user shall maintain Products in strict confidence and shall not: [...] (d) disclose, publish or otherwise make publicly available any benchmark, performance or comparison tests that End User runs (or has run on its behalf by a third party) on the Products;

I trust you can think through the implications of this restriction on your own. It certainly adds additional color to the recent public spat between NSS and Palo Alto Networks.

A lot of noise has been made recently about some nude celebrity photos that have made their way around the Internet, such as those of Jennifer Lawrence. Initially it was thought to have been a compromise of Apple's iCloud service, but, really it was because Apple failed to properly rate-limit an API that allowed rapid brute-force guessing of passwords. Apple claims it was a targeted attack on these specific users and tells users to enable two-factor authentication and use a strong password.

Getting someone's iCloud credentials, either by brute force or guessing the answers to security questions, is all well and good. It certainly explains these photos were acquired and posted. It doesn't adequately explain how supposedly deleted photos from a celebrity's iPhone were posted online. 

Moti Sagey, whom I work with at Check Point, actually came up with the idea of how this might have happened. He discovered a tool called dr.fone that can "recover data from iOS devices, iCloud backup, and iTunes backup." Curious, Moti and I downloaded and installed the tool, attempting to recover data from our own iCloud backups. Lo and behold: there they were. Up to 3 iCloud backups for each device we own. 

A note about the way iCloud backups work: they will only happen from your iOS device if the device is on WiFi and is plugged in--usually at night. They could also be triggered manually, but let's face it: how often do regular people do that? Almost never. 

While the majority of my devices have 3 consecutive days of backups, one in particular had quite a gap in their backup dates. Which, for most celebrity-types, may be pretty common. Movie stars are like normal people: they rarely charge their phone and are rarely on WiFi outside of their own home. Given how busy some celebrities are, it doesn't seem inconceivable that they could go months without being at home on their WiFi and charging their iPhone at the same time.

What's in an iCloud backup, you ask? Quite a lot, as it turns out, including the camera roll:

A popular celebrity with easy to guess iCloud username and password plus a rarely backed up iPhone plus a tool like dr.fone equals access to potentially naughty photos the celebrity thought were deleted months ago. Voila!

Of course, this is just the tip of the iceberg of what criminals do in order to get illicit photos and videos from celebrities, but let's be honest: why go through the trouble of social engineering, installing illicit remote access tools, and stealing authentication keys when you can take the much easier route of brute force guessing a weak password and/or easily answer their password reset questions?

People are blaming Apple here, when, last we checked, you had to opt-in to iCloud backups. Even if you choose to opt-in, you can choose what elements of your phone to back up. The fact that Apple keeps a few of them on-hand is, in fact, a good thing. Good backups are just as important as choosing a strong password or using two-factor authentication when its available. While there is some inherent risk in backing stuff up to the cloud, doing so is a greater good than never backing up your device, which is what would happen for most people without iCloud doing it for them.

Since it appears iCloud only keeps the last 3 device backups, ensuring that iCloud backs up your device regularly will also reduce the risk someone might "recover" an old nude photo you thought you deleted. Of course, if you want to reduce the risk even further, don't take the photos with your mobile phone, or at least don't do it with the default camera application. Use an ephemeral messaging app like Wickr or Telegram, at least until phones get an incognito mode similar to browsers.

Authorship Note: If Posthaven would let me mark more than one author, I would credit Moti Sagey as one. Especially since this was his idea, which I put a few refinements on.

Back when I worked for the Nokia TAC support Check Point Security Gateways, I actually ran a Nokia appliance as the headend for my network. One could call it "keeping my skills sharp for my daily work." For a variety of reasons, I discontinued this practice and instead opted for more traditional home routers that I could install DD-WRT on.

When I started working for Check Point, I alternated between either running a DD-WRT router or using a UTM-1 EDGE appliance. When Check Point announced the SG80 (and later the 1100 series appliance), I moved to using those. 

And while I probably could continue to use the 1100 series appliance for some time, I finally got my hands on a 2200 Appliance and decided that it should be the headend to my home Internet connection. I wanted to be able to use all the latest Check Point features and get my hands a bit dirtier than I have in a while. What better way to do that than with my family, which is very good at generating lots of real-world, Internet-bound traffic :)

DHCP and DNS Server on the Gateway

Every gateway I've used in the last several years had a DHCP server. In most cases, it also included a DNS server. Gaia does have a built-in DHCP server, and should you use it, should make sure to configure your rulebase to permit the traffic correctly as described in sk98839.

That said, it's not currently possible to make the Gaia DHCP server authoritative for a given subnet without manually hacking the dhcpd.conf file (see sk92436 and sk101525). I did this, which worked.

I still wanted a local DNS server too where I could define local DNS entries for items on my subnet. I could have easily used dnsmasq on one of the many DD-WRT routers I still have on my network as WiFi access points, but I wanted it on the Security Gateway, just like I had it on my 1100, UTM-1 EDGE, and others. I poked around a bit on my Gaia R77.20 installation and discovered that it too had dnsmasq.

I want to be clear: dnsmasq is not formally supported on Gaia. That said, it appears to work quite nicely. The configuration file /etc/dnsmasq.conf is not managed through Gaia and, as far as I know, it should survive upgrades and the like and not be affected by anything you do in Gaia, except maybe enable the DHCP server in Gaia, which obviously do you don't want to do if you're going to use dnsmasq for this.

The dnsmasq in Gaia appears to take most of the common configuration options, so I set about configuring it as described in the dnsmasq documentation. Starting it up is pretty simple: a service dnsmasq start from Expert mode gets it going.

But, of course, that only works until the next reboot. Now I could have hacked /etc/rc.local or something like that, but I know how to make Gaia's configuration system actually start this process for me, monitor it, and restart it if it dies.  

[Expert@animal:0]# dbset process:dnsmasq t
[Expert@animal:0]# dbset process:dnsmasq:path /usr/sbin
[Expert@animal:0]# dbset process:dnsmasq:runlevel 3
[Expert@animal:0]# dbset :save

I guess all those years of supporting IPSO paid off :) If you want to turn off dnsmasq in the future, issue the following commands from Expert mode:

[Expert@animal:0]# dbset process:dnsmasq
[Expert@animal:0]# dbset :save

One other thing you might want to do if you're running a local DNS server is have your Security Gateway actually use it. The problem is: dhclient (which obtains settings from a DHCP server) actually overwrites the configuration in Gaia for the DNS domain and the DNS servers to use. You should be able to prevent this by changing the /etc/dhclient.conf configuration file not to use this information.

Specifically, it means changing this line:

request ntp-servers, subnet-mask, host-name, routers, domain-name-servers, domain-name;

to this:

request ntp-servers, subnet-mask, host-name, routers;

DHCP and the Default Route on the Internet Interface

When I put the 2200 in my network, I noticed that Gaia was not accepting the default route from Comcast's DHCP server. To get my home network oeprational, I manually set the default route to my Internet-facing interface, which seemed to work for a time.

One persistent issue I was having, that I initially was unable to figure out, was that it would take some time in order for sites to load. I wasn't quite sure why, but it was only happening occasionally and not something I could easily track down. I ignored it for a couple of weeks, but last night the problem really came to a head.

I thought Comcast was having a problem. I even went so far as to call Comcast and they said there was some signal problem on their end. The problem still persisted. The only thing "different" was that we had visitors who brought a laptop with him and was playing League of Legends with my son.

I tried a number of things, and couldn't figure out what was going on. I did start seeing some peculiar traffic on my external interface, and decided on a fluke to check my arp table. Rather than the 20-30 entries I'm used to seeing, I was seeing over 700 entries. For IPs on the Internet. That didn't look right. And sure enough, I was seeing arp whois requests for Internet IPs coming from my gateway.

Arping for every IP you access on the Internet is not terribly efficient. Comcast was, understandably, not responding to these ARP requests quickly, though they eventually did. Furthermore, due to the transient nature of ARP entries, they were timing out every minute or so, which meant that as long as I was actively passing traffic, connections to websites worked great. When I stopped and reconnected? It would take forever to connect.

Which brings me back to the original problem I had: getting Gaia to take a default route from a DHCP server. Turns out, there's an option in Gaia called "kernel routes" that you need to enable to make this work, and enabling it is pretty straightforward. If I had search SecureKnowledge earlier, I probably could have saved myself learning (and documenting) this lesson. It is documented in sk95869.

Now the Internet at home is working great once  No delays in loading websites at all. At least ones caused by configuration errors on my part :)

Dynamic IP Gateway Security Policy

One of the challenges of using a Dynamic IP Gateway is that you don't know what the external IP is of your security gateway. Not knowing it could make it difficult to, say, make "port forwarding" rules or rules that relate to allowing traffic to the Security Gateway itself.

Turns out this is not an issue. When you check a gateway object as Dynamic IP as shown below, two Dynamic Objects come into play: LocalMachine (which always resolves to your gateway's external IP address) and LocalMachine_All_Interfaces which resolves to all of the IP addresses associated with your gateway. These dynamic objects are simply placeholder objects that get their actual definition updated on the local gateway. For these two objects in particular, this happens periodically without any action on the administrator's part.

Dynamic Objects have a couple of important limitations. Rules using Dynamic Objects must be installed on a specific installation target, i.e. you cannot simply leave the Install On field for the rule the default Policy Targets. Not doing so will cause a policy compilation error. The second issue is that rules with Dynamic Objects disable SecureXL termplates at that rule. This isn't a huge deal as I'm coming nowhere near the capacity of my 2200, and most of my traffic is hitting rules with acceleration, but it is a consideration you must make when constructing your rulebase. 

Another important limitation: Identity Awareness is not supported on Security Gateways that have a Dynamic IP. I do not believe this is an issue on the 600/1100 appliances, but it is for the 2200 and up. 

Final Thoughts

It's been a while since I posted something technical on my own blog about using Check Point. My role is such that I'm not always involved in the technical side of things like I was back when I wasn't quite Internet famous :)

In any case, it may be something I'll do again in the future. I'm sure that pieces of this article will also show up in SecureKnowledge--the supported parts of the above at least.

And, of course, these are my own thoughts and experiences. Corrections and feedback are welcome.

A blog post written by Greg Ferro, a.k.a. EtherealMind, had come to my attention on a review of ClusterXL in Check Point R77 based on the documentation. I'm familiar with Ferro from the Packet Pushers Podcast where he and Ethan Banks talk nerdy about networking. They occasionally dip into the security arena and when Check Point does come up, it's not in a positive light.

When I actually read Ferro's post about ClusterXL, I was not surprised that the end result was a negative review: not only of ClusterXL but of Check Point itself. What surprised me was how incomplete the review was, technically speaking, because I know how nerdy he can get.

Ferro only used one document for his paper review: the Check Point R77 Versions ClusterXL Admin Guide dated 28 July 2014. He could have easily asked the customer he supposedly did this review for to obtain additional documentation, such as the ClusterXL Advanced Technical Reference Guide in sk93306. Because he didn't have direct access to said documentation as an independent consultant, he chose not to use it.

What is ClusterXL anyway?

Even with the documentation Ferro chose to use, the more I read his Tech Notes Check Point Firewall ClusterXL in 2014 piece, the more I conclude his review was not very thorough. Clearly he read some of the documentation as there are quotes from it throughout the piece, but it's obvious he missed the part that defined what ClusterXL actually is:

ClusterXL is a Check Point software-based cluster solution for Security Gateway redundancy and Load Sharing. A ClusterXL Security Cluster contains identical Check Point Security Gateways.

• A High Availability Security Cluster ensures Security Gateway and VPN connection redundancy by providing transparent failover to a backup Security Gateway in the event of failure.
• A Load Sharing Security Cluster provides reliability and also increases performance, as all members are active

Ferro's definition of ClusterXL?

There seems to be two modes of active/active High Availability for CheckPoint Firewall one. Both of which are called ClusterXL.

Ferro then talks about the requirements for ClusterXL, which he gets mostly right: sync is a Layer 2 protocol, ClusterXL requires appropriate licenses (which every current Check Point appliance has, as do many current software-only firewall SKUs), and you should use NTP between members. Actually NTP is a good idea for reasons other than state sync (e.g. VPNs and SSL).

State Synchronization

Then Ferro comments on State Synchronization.

• by defaults all connection state is synchronised across the cluster.
• you can decide not to synchronise. Why ? Does this imply a performance issue in the devices, network, speed or other issues ? What would I not sync all state ?
• “Synchronization incurs a performance cost” – implies that CheckPoint performance remains a problem for “many customers

For the vast majority of Check Point customers, state sync does not have performance issues. Where issues have been observed is in environments where the connection rate is far higher than the throughput would suggest. You can reduce the load by disabling sync for services with connections that are transient in nature (e.g. DNS, short-lived HTTP connections) and can be done easily in SmartDashboard.

But that's not all he has to say about State Synchronization:

• Synchronisation Network requires elevated security profile, suggests CCP is insecure protocol, isolated network.
• Recommends using an isolated switch – this is stupidly impractical who the heck wants to waste money and power on a dedicated switch in the DMZ.
• Using a crossover cable is equally stupid and impractical. What are they thinking ?
• Appears that sync is ONLY supported on the lowest VLAN ID on an interface.

The Synchronization Network requires an elevated security profile: yes, this is stated in the documentation and is no different than every other major vendor. A dedicated switch or a crossover cable allows this elevated security profile to be maintained better than a dedicated VLAN, which by the way, if you want to use a dedicated VLAN, yes, this is supported; it's a common configuration in customer networks. As to whether a dedicated switch or crossover cable is "stupidly impractical," customers have asked for support for all of these options, which is why they are tested and supported options (and thus listed in the documentation as such).

Ferro's next complaint? You can't use more than one synchronization interface. This actually used to be a supported configuration once upon a time but in recent versions, the supported approach is to let the operating system handle interface redundancy via bonded interfaces (i.e. with LACP). I asked in the comments of the post why he felt using LACP was an inferior solution to multiple independent synchronization interfaces. No response.

ClusterXL Load Sharing

Ferro never once explicitly mentions the sort of configuration he was using as the basis for his paper review of ClusterXL. Based on the fact he only briefly mentions High Availability (or Active/Passive, as he said) and spends the majority of his time talking about Load Sharing (Active/Active), I can only assume that is the configuration he's most interested in, ignoring the fact that High Availability configurations are the far more common customer configuration.

Of the two ClusterXL modes that relate to Load Sharing, Ferro only talks about one: Load Sharing Multicast. This mode, according to Ferro, "should be avoided at all costs" because it involves implementing workarounds that are, in his words, "complex and hard to maintain/operate." These workarounds, static multicast mappings and/or disabling IGMP, are necessary because some networking equipment out there does not properly support RFCs or has issues with frame replication (needed for Multicast). 

ClusterXL has a mode that requires none of these workarounds: Load Sharing Unicast mode. This mode was only mentioned once, but no analysis of this mode was offered by Ferro.

Ferro's opinion about VPN and Load Sharing is that "there are many conditional statements about VPN connections" and "would have little confidence in running Load Sharing Multicast Mode and running VPN services on the same physical unit." I couldn't find the conditional statements he was referring to, but what I did find was one particular statement in relation to third party peers and load sharing (repeated in a couple of different ways):

The third-party peers, lacking the ability to store more than one set of SAs, cannot negotiate a VPN tunnel with multiple cluster members, and therefore the cluster member cannot complete the routing transaction.

The problem isn't Check Point, per-se, the problem is with interoperating with non-Check Point devices. The workaround: using Sticky Decision Function, which is listed in the documentation where this "conditional" statement is listed. This, along with the other so-called conditional statements he refers to are curiously absent from his paper review.

Performance and Cost

Ferro writes in several digs at Check Point's allegedly "punishingly expensive" products. In most of the competitive situations I've been engaged with during my tenure at Check Point, pricing compared to the competition is rarely a factor. That said, I have seen numerous situations where competitors have purposely specified lower-end (and thus cheaper) solutions that, in reality, will not meet the customer's stated requirements while maintaining the same security posture.

Ferro also states "forwarding performance is and price/performance is very poor" which is provably false. Independent third parties have verified Check Point's performance and pricing and have found Check Point in-line with the competition. It's also something being validated internally at Check Point on an ongoing basis.

Ferro is correct that performance decreases when additional security features are enabled, however this phenomenon is by no means unique to Check Point products. The expected performance for a given Check Point appliance is listed right on the data sheet along with the unrealistic "lab" numbers every networking vendor since the dawn of time has included on their datasheets. Unlike Check Point's competitors, the "real world" performance is evaluated with real-world traffic flows, multiple software blades (security functions), a real-world rulebase, and logging/NAT enabled. Your local Check Point account team can provide more refined appliance recommendations based on your specific requirements.

In Conclusion

As noted before, it's not really clear what use cases Ferro was evaluating ClusterXL against. Even his conclusion is non-specific about this:

Check Point doesn’t have strong solution for clustering and the weak documentation suggests that it’s not fit for critical use cases.

Perhaps Ferro does have a use case that ClusterXL would not be a good fit for. I'd love to know what those "critical use cases" are in more detail so his conclusion could be evaluated more thoroughly. 

Ferro's conclusion is based on an incomplete review of the available information. Only one document was reviewed when others exist, and I question how thoroughly even the one document was reviewed. Only one of the load sharing modes of ClusterXL was reviewed in any detail and he completely ignores the other mode as well as high availability configurations. Ferro also ignores the fact that practically all of Check Point's 100,000+ customers use ClusterXL successfully in their networks, which includes 100% of the Fortune 100 and Global 100 customers across many different industry verticals.

In short: Ferro's review of ClusterXL is incomplete and his conclusion is not supported by the facts.

Disclaimer: I work for Check Point Software Technologies. That said, the above are my own thoughts.

I decided that I'm going to let the phoneboy.net URL expire when it comes up for renewal in a couple of months. As a result of that, the URL for this blog (such that it is) will change to http://securitytheater.phoneboy.com

Please update your RSS readers, bookmarks, and the like.

Back when I first got into IT and just started working with FireWall-1, Pointcast was a thing. For those who weren't around back in the mid to late 1990s, Pointcast had a very popular screensaver that displayed news and other information delivered periodically over the Internet to PCs. The problem was: it used an excessive amount of bandwidth on corporate networks, especially if more than a couple of people used it.

The result was, of course, corporations wanted to block access to Pointcast. The problem: how to do it. All we had in the mid 1990s was the traditional firewall which could control access based on IP and port. So we should be able to block the port or IPs it communicates with, right?

Pointcast used good old HTTP. Even back then, no one in their right mind would block HTTP. Of course, everything uses HTTP or HTTPS to communicate these days, and with a traditional firewall with the ability to control traffic only by IP or port, leaving HTTP or HTTPS wide open is tantamount to leaving the barn door open. 

Pointcast didn't exactly publish their list of servers, but users of the PhoneBoy FireWall-1 FAQ contributed a list of IPs plus a couple of other clever solutions to the problem, which I've made available after the break if you're curious.

Of course, with things like content delivery networks, Amazon Web Services, and a host of other ways to serve up an application to users that are available today, attempting to control access to these applications merely by port and IP address is crazy. 

Fortunately, there are a number of solutions to this problem. Check Point's solution is the Application Control Software Blade, which can allow/block access to an application regardless of the ports and destination IP users, and even limit the bandwidth these applications use. New applications or changes to existing applications are made available to the gateway periodically so you can see that you're users are using it and, when it kills you bandwidth or worse, you can block it. 

If only tools like App Control were available back in the day, security admins could have spent more time on more important issues rather than figuring out how to block Pointcast and other applications and I would have a few less FAQ entries on "how do I block X application."

Read more »

From Chris Hoff's (a.k.a. Beaker) NGFW = No Good For Workloads:

NGFW, as defined, is a campus and branch solution. Campus and Branch NGFW solves the “inside-out” problem — applying policy from a number of known/identified users on the “inside” to a potentially infinite number of applications and services “outside” the firewall, generally connected to the Internet. They function generally as forward proxies with various network insertion strategies.

That is, as Beaker continues, very different from what a Data Center firewall needs to do:

Data Center NGFW is the inverse of the “inside-out” problem.  They solve the “outside-in” problem; applying policy from a potentially infinite number of unknown (or potentially unknown) users/clients on the “outside” to a nominally diminutive number of well-known applications and services “inside” the firewall that are exposed generally to the Internet.  They function generally as reverse proxies with various network insertion strategies.

In other words, we're not always sure who is coming in, but we know what they are going to and (hopefully) what applications and services they are going to connect to. 

What kinds of protection do you need in these scenarios? Usually very different. Can every next generation firewall provide just the right protection? 

First, let's take a step back and realize that the Data Center itself is very different from what it used to be a decade or two ago. Whereas we started with a number of servers hosting resources in one or two physical locations with users mostly in known physical locations, we now potentially have services, data, and users all over the place, with a mix of physical and virtual servers where traditional methods of segmentation and protection are not practical. 

The "core" of the enterprise network--where all the necessary resources ultimately connect together--is quickly becoming the Internet itself. How do you protect your resources in this reality?

We go back to one of the fundamental tenets of information security, our old friend segmentation. This means grouping together resources with like function and like information confidentiality levels, placing a enforcement point at the ingress/egress point where you can enforce the appropriate access control policy. The goal for that enforcement point? Let the authorized stuff in and keep the unauthorized and bad stuff out. 

Of course with virtualization, end user PCs, and mobile devices, the boundaries become more difficult to apply but with virtualized security solutions, integrated endpoint security on the end user PCs, trusted channels (VPNs), and secure containers on mobile devices, more is possible than you think. Check Point and other companies have various solutions for this. 

Once the network is segmented and enforcement points are in place, then you can decide what protections and policies should be applied. In some cases, like on User Segments, you want lots of protection as users could go anywhere on the Internet and unknowingly bring in some malware to run amok in your network or send company secrets to their Gmail account. For your data center? Maybe you just want to make sure authorized users can reach specific applications and you want to sanity check the traffic to make sure it's not malicious. Or maybe you just need a simple port-based firewall with low latency for a given app.

The idea of putting a firewall as the core of your network--especially a next generation one-- is silly, as Beaker rightfully points out. Really, your core should be a transit network with enforcement points--those things we typically call firewalls--at the ingress point of the various network segments. This way, just the right policy and just the right protections can be applied without applying them to traffic that doesn't need it. 

This is where I think Check Point's portfolio shines. In the Security Gateway space, the Software Blades architecture is flexible enough to allow you to be very granular about what protections are applied to a specific enforcement point, whether a physical gateway, or a virtual one either in a Check Point chassis (e.g. VSX) or in a VMware or Amazon Web Services environment. This means you can scan a random MS Word document from the Internet for malware on one gateway close to the users while not impeding the flow of traffic in and out of your Data Center that flows through a different Security Gateway. And yes, if you have a 5 microsecond transmission requirement, Check Point has a solution for that with the Security Acceleration Module in the 21000 series of appliances. 

Does an NGFW solve every problem? No, and anyone that tells you it will is flat out wrong. It's not always the right tool for the job, as Beaker points out:

Show me how a forward-proxy optimized [Campus & Branch] NGFW deals with a DDoS attack (assuming the pipe isn’t flooded in the first place.)  Show me how a forward-proxy optimized C&B NGFW deals with application level attacks manipulating business logic and webapp attack vectors across known-good or unknown inputs.

Likewise, business logic and webapp attack vectors are outside of the wheelhouse of all NGFWs. You still need to properly secure your web applications, even with an NGFW in place. In addition, there are dedicated, Web Application Firewalls for this purpose and if you've properly segmented your network, you can make sure only those resources are protected by them.

At the end of the day, a Next Generation Firewall, whether it is from Check Point or someone else, is not a panacea. It can be a powerful tool, but like all tools, it needs to be applied properly as part of a comprehensive security strategy that begins with proper segmentation and a well-defined policy. From there, you can apply just the right protection to just the right resources.

Disclaimer: It should be obvious from my last post I work for Check Point, but this is my own opinion.