Our Infrastructure
Many of our competitors are just "private label brands" on some other companies platform. PizzaCloud owns and managed our infrastructure and platform, giving us total control over features and reliability. In addition, more and more phone companies operate solely in the cloud - frequently in Amazon cloud services. But cloud services fail...Amazon in particular has had numerous outages, some national. We have all seen the headlines like this:
Prolonged AWS outage takes down a big chunk of the internet
AWS has been experiencing an outage for hours: Nov 25th 2020
We maintain our own physical systems in two data centers (Columbus, Ohio and Trenton Tennessee), plus use cloud services in two Google Cloud Engine (GCE) data centers in California and South Carolina. Why Google? Because Google's GCE cloud services are the most reliable in the world. The specific part of Google we use - GCE virtual machines have over 99.999% uptime. In fact, they have had less than two minutes of down time in the last three years, giving them 99.9999%.
Why also have physical data centers? Because nothing is perfect. Nothing is 100% reliable. We do not like to have "all our eggs in one basket".
Every customer has a primary and failover data center assigned. If we lose a server, router or have any other issue at your primary data center the phones snap over to the backup in seconds.
This gives us the best of both worlds - the flexibility of cloud services with the tried and true reliability of physical servers.
In general, customers on the west coast are primary in the LA google data center, failing over to our physical data center in Trenton. Customers in the south East are primary in the Google SC data center, failing over to our physical data center in Columbus Oh, and everyone else is either primary Columbus/failover to Trenton or the reverse.
Our Columbus data center is a "tier 0" data center with multiple power feeds, redundant battery backups, multiple 5Mwatt generators, 6 fiber Internet connections etc. Virtual tour here: https://expedient.com/data-centers/columbus-upperarlington/
So - what exactly does "physical servers" mean? Well, we run a multi-server high-availability virtualization platform. Meaning that if we lose a physical server the virtual machines running on it pop up on the other servers. Servers are connected by multiple 100gb networks. We are in the process of upgrading both physical data centers to about twice the performance, memory and storage shown below. This is scheduled for completion first week of Jan 2021.
Prolonged AWS outage takes down a big chunk of the internet
AWS has been experiencing an outage for hours: Nov 25th 2020
We maintain our own physical systems in two data centers (Columbus, Ohio and Trenton Tennessee), plus use cloud services in two Google Cloud Engine (GCE) data centers in California and South Carolina. Why Google? Because Google's GCE cloud services are the most reliable in the world. The specific part of Google we use - GCE virtual machines have over 99.999% uptime. In fact, they have had less than two minutes of down time in the last three years, giving them 99.9999%.
Why also have physical data centers? Because nothing is perfect. Nothing is 100% reliable. We do not like to have "all our eggs in one basket".
Every customer has a primary and failover data center assigned. If we lose a server, router or have any other issue at your primary data center the phones snap over to the backup in seconds.
This gives us the best of both worlds - the flexibility of cloud services with the tried and true reliability of physical servers.
In general, customers on the west coast are primary in the LA google data center, failing over to our physical data center in Trenton. Customers in the south East are primary in the Google SC data center, failing over to our physical data center in Columbus Oh, and everyone else is either primary Columbus/failover to Trenton or the reverse.
Our Columbus data center is a "tier 0" data center with multiple power feeds, redundant battery backups, multiple 5Mwatt generators, 6 fiber Internet connections etc. Virtual tour here: https://expedient.com/data-centers/columbus-upperarlington/
So - what exactly does "physical servers" mean? Well, we run a multi-server high-availability virtualization platform. Meaning that if we lose a physical server the virtual machines running on it pop up on the other servers. Servers are connected by multiple 100gb networks. We are in the process of upgrading both physical data centers to about twice the performance, memory and storage shown below. This is scheduled for completion first week of Jan 2021.
Our primary border routers are high availability pairs of Peplink Balance SDX-Pro. 24gps throughput.
Our unique phone system is a multi-data center high availability cluster phone platform. Any given phone can connect to any node in any of the four data centers. Calls from carriers can come into any data center. of course we normally have all phones for given customer set to one switch node in one data center, failing over to a second data center.
Then we have multiple such clusters (again, the "not all eggs in one basket" mode).
How much traffic do we handle? Below is a typical recent Friday evening snapshot of one of our four phone system clusters. PizzaCloud is one of two "brands" we operate. The other provides phone service to government offices and E911 centers in several states.
Our unique phone system is a multi-data center high availability cluster phone platform. Any given phone can connect to any node in any of the four data centers. Calls from carriers can come into any data center. of course we normally have all phones for given customer set to one switch node in one data center, failing over to a second data center.
Then we have multiple such clusters (again, the "not all eggs in one basket" mode).
How much traffic do we handle? Below is a typical recent Friday evening snapshot of one of our four phone system clusters. PizzaCloud is one of two "brands" we operate. The other provides phone service to government offices and E911 centers in several states.
Does all that mean "My phones will NEVER go down?".
Frankly - no, it does not mean that. Anyone who promises 100% uptime is lying to you. AT&T has had regional and nationwide outages. Verizon has had regional and nationwide outages. Comcast has had many such outages. If the largest phone companies in the world cannot promise 100% uptime, who can?
Same on the cellular backup Internet. Your primary Internet goes down, the cellular backup kicks in. But what if your area had a terrible ice storm, tornado or similar and the cell towers near you are down as well?
Bottom line - we have layer upon layer of redundancy and backups, but that only takes us to 99.999 reliability. Which in fact is far better that "plain old telephone service", but again - anyone who promises perfection is lying.
Frankly - no, it does not mean that. Anyone who promises 100% uptime is lying to you. AT&T has had regional and nationwide outages. Verizon has had regional and nationwide outages. Comcast has had many such outages. If the largest phone companies in the world cannot promise 100% uptime, who can?
Same on the cellular backup Internet. Your primary Internet goes down, the cellular backup kicks in. But what if your area had a terrible ice storm, tornado or similar and the cell towers near you are down as well?
Bottom line - we have layer upon layer of redundancy and backups, but that only takes us to 99.999 reliability. Which in fact is far better that "plain old telephone service", but again - anyone who promises perfection is lying.