One should never underestimate the importance of platform stability and system “uptime” to an FX brokerage. These are critical aspects of the business that should never be overlooked and it is therefore vital that you make the right choices when considering someone to host your company’s server(s). By entrusting hosting to a third party, you are effectively dependent on that provider’s expertise, reliability and trustworthiness. Remember that, should they fail, every minute of downtime could cost you a considerable amount of money and, perhaps even worse, seriously damage your brand credibility.
Just think about this type of nightmare scenario; clients unable to load their platforms, unable to get out of trades, unable to view their market exposure and, of course, prevented from capturing market movements. Now imagine that is a recurring nightmare, something that happens repetitively. Your brokerage will become known for instability and unreliability and your revenues will suffer as a result. It therefore goes without saying that finding a stable and reliable hosting partner should be a priority when looking to rapidly grow a business. I cringe when I hear of brokerages trying to negotiate cheap hosting fees from unproven providers; after all, just one outage could cost them an entire year’s worth of hosting fees and much, much more. I guess the saying holds true, “penny wise and pound foolish”.
What are the most important evaluation criteria for the MT4 hosting provider?
1. Credibility and history of uptime.
You want, and need, your hosting partner to be credible with a history of reliability and to have a presence within the datacenter facilities that host major, top tier financial companies. For an FX brokerage, presence in New York’s NY4 is a must. The majority of firms providing hosting services tend to go through two or three layers of managed hosting providers to get their servers into NY4. Think about it; being two, three or even more steps removed from the actual source/vendor puts you in a position where issues cannot be resolved in a timely manner and where changes cannot be implemented within a desired timeframe.
Fortex, unlike many hosting providers, has its own cage in NY4 and (Hong Kong’s) HK3 and has built up a very impressive infrastructure that is able to handle significant load. We actually upgraded our NY4 infrastructure at the end of this past summer by deploying two new enterprise grade Fortinet firewalls to increase our network bandwidth throughput, allowing us to stream even more market data to our clients (increased number of updates per second).
2. Redundancy, redundancy and redundancy
Redundancy is very high on the list of priorities as this greatly helps to mitigate the risk for the brokerage company. You need to ensure that “hot backup” is in place for your mission critical software and also make certain that everything has built-in redundancy: use at least 2 CPUs, multiple hard drives (with RAID controllers), multiple ISPs, multiple network providers and multiple power sources. The Fortex network always utilizes dual firewalls, dual routers, dual switches, dual paths of connectivity and multiple Internet “backbone” carriers as ISPs; so in reality hot-hot back-up for the database and trading systems.
This is one area of your business where you shouldn’t be looking for savings, but rather for safety.
At Fortex, we understand how critical redundancy is for FX brokerages. Not only do we have five ISPs in our NY4 cage, but we also have multiple network paths. We host our clients on top-of-the-line dedicated hardware to ensure that every eventuality is covered where possible. Fortex network is architected in such a way that there no single point of failure.
Most brokers cannot estimate the amount of bandwidth they will need for their business so, many times, they simply go with whatever the hosting provider offers in their basic package. The big problem with this is that many hosting providers don’t understand the type of business that you are in and the type of software that you are planning to utilize. They therefore cannot accurately advise you on the right amount of bandwidth to subscribe to. With bandwidth, it’s my opinion that there are two important things to understand:
1. Make sure your bandwidth is “burstable”. If you’ve underestimated your bandwidth needs then, at least, make sure it is burstable. This will prevent downtime and connection issues even if you go over your allotted bandwidth. In the NY4 and HK3 datacenters, Fortex Internet bandwidth is burstable up to 2Gbps.
2. Also, there is often a misconception that all bandwidth offerings are created equal. This cannot be further from the truth. Most of the players out there use shared bandwidth that is throttled down. Fortex always uses multiple ISP (Internet Service Providers) at the same time. For example, for NY4 we have 5 ISPs, all of whom are Internet backbone carriers, such as Level 3, Verizon and Sprint. This set-up is not just for redundancy but, also to give us the largest global footprint on the Internet.
We locate the users Internet traffic dynamically and route their traffic through the ISP with the most efficient and stable routes. Fortex’s100mbps bandwidth is really very different from a tier2 or tier3 provider’s 100Mbps bandwidth where those providers are effectively re-selling a bundled line from a backbone carrier. In addition, all of our network gears and routing paths are paired as twins providing hot-hot redundancy in real-time. For example, for the Fortex production, we always run on two core switches, two edge switches and four firewalls at any given moment.
Typically, the discount wholesaler of bandwidth can lease a 1Gbp line and sell 50Mbps of line to fifty, not twenty end users knowing that they won’t all be using their allocated bandwidth at the same time. If they all do, then the peak usage hits 2.5Gbp and multiple packets will be dropped. In fact, often by the time the bandwidth gets to the users it has already been resold two to three layers down. This is why a 5Mbp home DSL or 20Mbps home Cable Internet are nothing when compared to a commercially leased line.
When you surf the web, download a big file or watch video online, your 20Mbps line might drop 25% of the packets and retransmit without your noticing but in reality you don’t need those to be real-time. Trading is a completely different story; every drop creates a delay. On Fortex’s internet lines and cross-connect, you can run 20,000 pings and not drop a packet but for many of those other internet services, you will see a 5% 10% or, sometimes, a 20% drop and this will ultimately result in massive trading disruption.
What extra services do you get by working with hosting providers that understand the FX business?
There is a definite advantage to be gained when your hosting vendor fully understands the business that you are in. At Fortex, our IT Team proactively monitors network capacity and latency to deliver the highest possible system uptime to our customers. We also assist our clients with software installation, data migration and will even make occasional recommendations in relation to system optimization.
Clients working with Fortex can take advantage of our standby Network Engineering team who have top notch expertise in working with all major exchanges and ECNs.
With 10+ years of experience in dealing with MT4 under our belt, our global support team delivers dependable and timely assistance and expert service that you can always rely on.
Rest assured that we have the knowhow and discipline to run large scale, mission critical systems, it is in our DNA. Our headquarters, and main operations hub, is in Silicon Valley California and, in everything we do, we remain focused on redundancy, recovery and scale.
Can you expect ZERO problems from your hosting provider?
Issues are unfortunately inevitable and occasional problems will happen, no matter how good of a provider you are dealing with. Your relationship with the hosting provider will strengthen if, instead of pointing fingers and assigning blame, you both work together to understand the situation and then take the necessary steps to ensure that the same problem doesn’t happen again in the future.