Access Layer Switch Refresh

When do you start thinking about your access layer switch refresh

The timing for planning a network access switch refresh depends on various factors, including the current state of your network infrastructure, technology advancements, budget cycles, and business requirements. Here are some considerations to help you determine when to start planning for a network access switch refresh:

Equipment Lifecycle: Network switches typically have a lifecycle, and manufacturers may provide end-of-life (EOL) and end-of-support (EOS) dates for their products. It’s advisable to plan for a refresh before reaching these dates to ensure you receive continued support, software updates, and security patches.

Technological Advancements: Stay informed about advancements in networking technologies. If there are significant improvements, such as faster data rates, enhanced security features, or improved energy efficiency, it may be worth planning a refresh to take advantage of these benefits.

Performance and Scalability: Assess the performance of your current network switches and evaluate whether they meet your current and future needs. If your network is experiencing performance issues or lacks the scalability required for future growth, it might be time to plan for a refresh.

Security Concerns: Security is a critical consideration. If your existing switches lack the latest security features or are no longer receiving updates, it’s crucial to plan for a refresh to maintain a secure network infrastructure.

Budget Cycle: Align your network switch refresh with your organization’s budget cycle. Planning ahead allows you to allocate funds and resources for the new equipment, ensuring a smoother transition.

Business Requirements: Consider any upcoming changes in your business that may impact your network requirements. For example, if you’re planning to expand your operations, introduce new applications, or adopt emerging technologies, these factors may influence the timing of your network switch refresh.

Evaluation and Testing: Before implementing a refresh, it’s essential to evaluate and test potential new switch models to ensure compatibility with your existing infrastructure and to validate their performance under your specific network conditions.

In summary, it’s advisable to start planning for a network access switch refresh well in advance to ensure a smooth transition, mitigate risks, and take advantage of the latest technological advancements. Regularly assess the state of your network infrastructure and stay informed about industry trends to make informed decisions about when to refresh your network switches.

The end goal is to tidy up the cabinets and replace your access switches

Our goal as experts in the field is to not only replace the access switches with the latest technology but to also make sure that the end result looks much better and is a lot easier to manage. If you look at the left hand side picture of the cabinet it is clear to see that over the years it has become very messy with loads of cables that are way too long. The cable management is hardly used and it is even difficult to close the cabinet door.

On the right hand side you can see exactly the same cabinet with new switches but what stands out here is the cabling. Each patch panel is split into two halves to make it look tidy and the amount of cables between the racks is minimised. All cables are nice and tight together with a lot of velcro. Any redundant or unused patch leads are removed from the rack.

At EDNX we treat cabling like work of art and take pride in making sure it looks spotless in the end of the process. In the next paragraph we will unveil some trade secrets to explain how you can tidy up the cabinet this way while making sure that any outage or mistakes are avoided. In order to do this we have to split the physical aspects from the logical constructs.

Start by performing the initial cabinet audit and tidy up

First thing that needs to be done is the initial cabinet audit for power, space and connectivity. We need to assess if the new switches can be collocated and powered up or is longer outage required to allow un-racking of old devices. Another very important aspect of the initial scoping should be disconnecting all unused ports from the patch panels. At EDNX we have our own bespoke script that can be run to generate a spreadsheet showing all port configurations and highlighting all of the ports that have been down for a given amount of time like 12 weeks. This allows us to quickly remove a lot of unused ports clearing the cabinet as well as asses how many new ports would be required. For example you may have a stack of 4 x 48 port switches but based on port usage analysis only 60% of ports have been active in the last 12 weeks which could reduce the number of new switches that are required to 3. Our script has a graphical user interface that allows us to run the same assessment for any number of weeks so if 12 weeks is not long enough we could include 40 weeks as a parameter.

The output from the script would be an Excel file that shows all different config permutations along with all ports in RED that haven’t been active for a given amount of time. All those ports could be disconnected even before the official outage window allowing a lot more space in the cabinet. All other ports have to be traced back to the patch panel as explained below.

Physical switch replacement doesn’t sound too complicated but how about the logical configuration?

Lets assume for now that this is not SDA network which makes the logical configuration easy. In SD-Access every port is the same and Cisco ISE dynamically allocates it to the correct VLAN or VRF. In the traditional environment each access switch could potentially have a number of VLANs randomly spread across all ports. Those could be end user laptops, access points, security cameras, printers and plethora of other wired devices that are more or less important to the business. This makes things quite a bit more complicated because we need to figure out which port on the patch panel corresponds to a given device to make sure we patch it to the same VLAN on the new switch. If the switch only has two or three VLANs it is possible that one of them would be what we call dominant configuration. We can then trace all of the remaining ports on the switch back to the patch panels and document it leaving the dominant configuration as everything else that remains on the patch panel.  One way or the other we have to end up with patch panel ports to VLAN mappings in order to plan configuration of the new switch. One example is shown on the drawing below where two switches are in the rack and each patch panel port is traced to the relevant switch port. The colours indicate different VLAN in this case orange – access point, blue – security camera and green for standard user port.

Change control and network implementation plans

Once we have all required information from the physical point of view as well as port to VLAN mappings on the new switch, the final step is to prepare Network Implementation Plan (NIP). This document could be presented to Change Approval Board (CAB) to assess any risk and schedule the migration date. Any well prepared NIP would include all low level information required to execute the change both from the physical and logical perspective. It would include well formatted diagrams, comments and a back out plan. From our experience it normally takes about 80-90% of time to plan the change while only 10-20% to implement it. Every network engineer should always remember that single line of configuration or one wrong port getting disconnected or placed in a wrong VLAN can cause a very serious outage. This is why spending enough time on meticulous planning is so important and cannot be overstated.

With hundreds of network changes we have conducted over the years we have mastered our implementation approach. Nevertheless we are still looking for marginal gains in the area of scripting and reducing risks for the customer.