Scott McIntyre’s Post

Efficient Link Load Balancing for Reliable Internet Connectivity

⁉ Question of the day: Are Fabric Interconnects switches or not... ⁉ FI's are purpose-built TOR devices to connect your UCS ecosystem to the rest of the network, and as such they ACT like switches in the following ways: ⚡ Provide network uplinks to the LAN ⚡ Provide Fibre Channel uplinks to storage ⚡ Facilitate inter-blade connectivity (no switching is done in the UCS chassis) However, they can be configured in 2 modes: 🖥 End Host Mode (default) - presents the FI as a host with multiple MAC addresses to the upstream switchd 🖥 Switching Mode - the FI operates as a limited functionality switch and subsequently requires spanning-tree / port-blocking End Host Mode is by far the most common. Without the need for STP there are less network disruptions. Additionally, the FI will learn all the MAC addresses of the blades within the system but does not need to learn devices on the network, because if it's not one of the servers then the frame is pushed out the uplink port, resulting in less load on the FI. No data is passed between the L1/L2 links between Fabric Interconnects, those are for management only. In switching mode, the FI's learn all MACs and do have switching capabilities, hence they MUST be configured for STP or else they can cause convergence issues. 💡 These modes exist for both Ethernet, and Fibre Channel within the Fabric Interconnect. 💡 Check the chart below for the forwarding rules used by Fabric Interconnects in End Host Mode! . . .

About 3 months ago, a new client at the company escalated an issue: both their LAN network and Wi-Fi were down. A team of network engineers, including myself, was dispatched to resolve it. I was excited 😁 —little did I know it would be a long day ahead! We arrived onsite bright and early, only to find their server room in chaos. The network rack was a tangled mess—cables running everywhere, left and right. Our first task was troubleshooting. My team lead asked me to ping Google’s server, but the ping returned the message "Request Timed Out". We ruled out the ISP, as the router had internet access. The real issue? The client's 3 switches. We had to trace and isolate each switch, and with the client having 3 floors, this wasn't an easy task. Using a network cable tracer, we started identifying where each cable was connected. And as we traced each one, we labeled it properly to improve future troubleshooting and applied proper cable management. After about 4 hours of tracing cables, I made an interesting discovery: one of the ethernet cable had both ends plugged into a Switch. That’s right—an IT personnel at the client site had created a Layer 2 loop, also known as a Switching Loop or Broadcast Storm. This type of loop occurs when multiple paths exist between switches. Normally, switches send out broadcast messages, but with a loop in place, the switch flooded the network with continuous broadcasts, causing severe downtime. Imagine coming to work and finding no internet access for four hours straight. For our client, this translated directly to lost productivity and revenue. To mitigate the issue, we ·        Removed the cable causing the loop. ·        Enabled Spanning Tree Protocol (STP) and Bridge Protocol Data Unit (BPDU) Guard on the switches to prevent future loops and ensure the fastest network path in the presence of redundant links. ·        Disabled all unused ports to prevent similar incidents in the future. ·        Educated the client’s IT personnel on what went wrong and how to avoid this in the future. Fast forward to today—we returned for a routine maintenance visit, and guess what? The network rack is in worse shape than it was three months ago! 😅

One of my Cybersisters shared her experience on how her team and she were about to successfully troubleshoot the cause a client’s LAN and WiFi’s downtime. Turned out to be a Switching loop problem.😅 This occurs when more than one path exists between the source and destination devices. In this case, the switch must have repeatedly sent broadcast messages, flooding the network and creating a broadcast storm. Read and digest. #networking #switches #cables #SpanningTreeProtocol

Ericsson launches Compact Packet Core in shadow of high profile outage Last week, Danish operator TDC Net suffered an outage that meant users couldn’t make calls, including to emergency services, for several hours. In a statement the day after the outage, the operator clearly fingered the likely guilty party as a technical error in its Ericsson core network software. Describing the event as a “significant operational disruption”, TDC Net said that “an unusually high volume” of call traffic had “overloaded part of the core network.” Ericsson, which also provides managed services to TDC, provided a statement to TDC that appeared to shoulder the blame. “Ericsson regrets the incident on Thursday in Denmark, which affected TDC NET’s mobile network. We are now working hard with our customer to identify the cause and will do our utmost to prevent it from happening again.” Now, Ericsson has launched a new, pre-configured version of its cloud-native core software, that it says will make the transition to cloud, and to 5G Core, easier for operators to handle and manage. https://lnkd.in/dNN4KgHk

Wi-Fi 7 offers significant advancements in speed, efficiency, and reliability, but the upgrade process may vary depending on the existing infrastructure, hardware compatibility, and organizational requirements. Understand key steps in the upgrade path.

Wi-Fi 7 offers significant advancements in speed, efficiency, and reliability, but the upgrade process may vary depending on the existing infrastructure, hardware compatibility, and organizational requirements. Understand key steps in the upgrade path.

Wi-Fi 7 offers significant advancements in speed, efficiency, and reliability, but the upgrade process may vary depending on the existing infrastructure, hardware compatibility, and organizational requirements. Understand key steps in the upgrade path.

Wi-Fi 7 offers significant advancements in speed, efficiency, and reliability, but the upgrade process may vary depending on the existing infrastructure, hardware compatibility, and organizational requirements. Understand key steps in the upgrade path.