4.6 Moving Forward: Designing for Reliability and Growth

Networks today have to handle way more devices than anyone imagined back in the 1980s. Phones, laptops, smart fridges, doorbell cameras, cars: pretty much everything needs an IP address. That reality pushed two big shifts in how networks are built. First, a new addressing system called IPv6 to keep up with demand. Second, careful design choices that keep networks running even when something breaks. This topic is about both of those moves.

Why IPv6 Exists

The internet was originally built on IPv4, which uses 32-bit addresses. That gives you around 4.3 billion possible addresses. Sounds like a lot until you realize the planet has way more than 4.3 billion connected devices. We've been running out of IPv4 addresses for years, and that shortage is called IPv4 address exhaustion.

IPv6 was designed to solve two problems at once:

• Keep up with the rapid growth in the number of devices joining the internet
• Deal with the anticipated depletion of available IPv4 addresses

The jump in address space is huge. IPv6 uses 128-bit addresses, which means the total number of possible addresses is $2^{128}$. That's around 340 undecillion addresses. In practice, this means every single device on the planet can have its own globally unique and routable IP address. No more sharing addresses through workarounds. Your phone, your laptop, your smart thermostat, each one can hold a real public address that routers across the internet can find directly.

IPv6 also moves data more efficiently than IPv4. The headers (the labels attached to each packet) are simpler, and the routing procedures routers use to pass packets along are streamlined. Less time spent processing each packet means faster, smoother traffic.

How IPv6 Addresses Are Written

An IPv6 address is 128 bits long. To make that readable, those bits get split into 8 hextets. Each hextet is a group of 16 bits, written in hexadecimal digits (0 to 9 and A to F), with a colon between each one.

A full IPv6 address looks like this:

</>Code
2001:0DB8:0000:0000:0000:0000:0000:1234

That's hard to read and even harder to type. So IPv6 lets you shorten addresses using two rules.

Rule 1: Drop Leading Zeros

Inside each hextet, you can remove zeros at the front. The hextet 000A becomes A. The hextet 01FF becomes 1FF. You cannot drop zeros from the middle or the end of a hextet, only from the front.

Rule 2: Collapse All-Zero Hextets with ::

If you have one or more hextets in a row that are all zeros, you can replace that whole run with a double colon ::. You can only use :: once in an address, because otherwise it would be unclear how many zero groups got replaced.

Shortening in Action

Here are the examples you should recognize:

Example 1:

</>Code
Full:      2003:AB00:CDEF:000A:0000:0000:0000:0001
Shortened: 2003:AB00:CDEF:A::1

The 000A dropped its leading zeros to become A. The four hextets of zeros collapsed into ::. The final 0001 dropped its leading zeros to become 1.

Example 2:

</>Code
Full:      2001:0DB8:0000:0000:0000:0000:0000:1234
Shortened: 2001:DB8::1234

0DB8 became DB8, the five all-zero hextets collapsed into ::, and 1234 stayed the same.

Example 3:

</>Code
Full:      FE80:0000:0000:0000:01FF:FE23:4567:890A
Shortened: FE80::1FF:FE23:4567:890A

Three zero hextets in the middle collapsed into ::, and 01FF lost its leading zero to become 1FF.

If you need to go the other way (expand a shortened address back to its full form), count how many hextets are shown, subtract from 8, and that's how many all-zero hextets the :: replaced.

Designing Networks for Availability and Reliability

A network is only useful when it actually works. Availability means the network is up and accessible when users need it. Reliability means it keeps performing correctly over time. The main strategy for both is redundancy: having a backup ready so that if the main (primary) system fails, a secondary option takes over without users losing service.

Common Redundancy Approaches

There are four big ones to know.

Backup power. If the electricity goes out, your network dies with it unless you have a plan. A UPS (uninterruptible power supply) is basically a big battery that kicks in instantly during a power loss. Generators handle longer outages by burning fuel to produce electricity. Together they keep servers, switches, and routers running through outages.

Secondary switches and routers. Adding extra switches and routers alongside the primary ones means that if one device fails, traffic can reroute through the backup. This is sometimes called hardware redundancy. A hospital, for example, can't afford for its network to drop just because one router died.

Backup internet connections. If your main ISP (internet service provider) has an outage, a secondary connection takes over. This is often a satellite link or a mobile (cellular) connection. It might be slower than the primary, but it keeps the business online during primary ISP outages.

Regular data backups. Copying critical data to backups on a regular basis protects against data loss. If a server crashes or ransomware locks up your files, you can restore from the backup. The key words here are regularly and critical: you need recent copies of the data that actually matters.

Monitoring and Alerting

Redundancy helps once something breaks, but you also want to know the moment something goes wrong. Network monitoring tools watch traffic, device health, and system logs around the clock. When they spot a failure or suspicious activity, they raise alerts so IT staff can respond quickly. Rapid detection shortens downtime and limits damage.

Monitoring also catches warning signs before full failures. A switch running hot or a link dropping packets occasionally might be about to fail completely. Catching that early lets you fix it on your own schedule instead of during an emergency.

Security Controls That Protect Uptime

Outages don't always come from hardware dying. Attacks and misconfigurations cause plenty of downtime too. Security controls can reduce the chance of those outages:

• Firewalls filter traffic to block malicious connections before they reach internal systems
• Intrusion prevention systems (IPS) actively detect and stop attacks in progress
• Access controls limit who can log in, change settings, or reach sensitive resources, which reduces both attacks and accidental misconfigurations

A locked-down network is a more available network. An attacker who can't get in can't knock your systems offline.

Patching and Documentation

Software has bugs and vulnerabilities. Vendors release updates and patches to fix them. Applying and documenting those patches is a core reliability practice. Patching regularly cuts the likelihood of failure caused by known bugs, and it closes security holes before attackers can use them.

Documenting the patches matters just as much as applying them. If an outage happens, having records of what was installed and when helps you figure out the cause faster. That shortens recovery time. Good documentation turns a mystery outage into a traceable problem.

Putting It Together

A reliable network stacks these methods on top of each other. A well-designed setup might have a UPS and generator for power, duplicate switches and routers, a backup cellular ISP, nightly backups of critical databases, 24/7 monitoring with alerts, a firewall and IPS at the edge, strict access controls, and a documented patching schedule. No single method covers everything. Layered together, they make the network tough enough to handle both expected growth and unexpected failures.