Upstream and Downstream QoS.īasic Troubleshooting/Connectivity. Supported Protocols, Specifications, and Certifications. Troubleshooting AP Mobility.Ĭlient Association Packet Flow.Ĭlient Utilities and Logging.ĪP Debugs and Show Commands. Primary, Secondary, and Tertiary Controllers. Mobility Role of the Controller to the Client.Ĭonfiguring Mobility Groups. Inter-Subnet Roaming/Layer 3 Mobility Events. Troubleshooting the AP Discovery and AP Join Process. Understanding the AP Discovery and AP Join Process. DHCP Bridging.Ĭonfigure the Switch for the WLC.ĪP Discovery and Join Process. LWAPP Layer 3 Transport Mode.ĭHCP Proxy Vs. Layer 2 and Layer 3 LWAPP Transport Modes of Operation.
Location Design Considerations.Ĭhapter 7 Deploying and Configuring the Wireless LAN Controller.Ĭonnecting the WLC to the Switch. Troubleshooting CAPWAP Session Establishment/AP Discovery and Join.ĬAPWAP Communication: Control and Data Encryption.ĬAPWAP Communication: Sequence Numbers and Retransmissions.ĬAPWAP-Data Packets Fragmentation.ĬAPWAP–MTU DISCOVERY and TCP-MSS Adjustment.ĬAPWAP-Data Binding and Payloads.ĬAPWAP-Control Binding and Payloads.ĭense AP Deployment Considerations. Manually Dissecting the Discovery Response.Ĭhapter 4 The CAPWAP Protocol.ĬAPWAP Session Establishment/AP Joining Process.
WLC Hardware and Software Requirements.ĭissecting the Discovery Response. Wireless LAN Controller Platforms.įunctionality Differences Between WLCs. Step 4: Analyzing the Data Collected About the Problem. Build quality is about the same as Supermicro's, too.Production Versus Nonproduction Outages. if your staff can hack the additional hardware QA running a Supermicro landscape entails they are IMHO a perfectly valid solution nowadays) for features you cannot use. Without the FI, you can't do any of those things, but you'll still likely pay a premium over HP or Dell (or Supermicro. There's also a bunch of nice other centralization things in that you can reach all the server's virtual KVM via the FI, and that you don't have to monitor each server for hardware health individually (the FI monitors the server hardware for you and writes issues to its event log, so you just monitor that) Rapid configuration and provisioning via templates (also, since we use blade chassis: less cabling mess) and the ability of the FI, in combination with the Palo NICs in the server, to emulate up to 64 virtual Ethernet and FC adapters, each with its own speed setting that can be configured freely between 1 Mbit and 10Gbit is basically one of the main points of UCS. We run a couple of FIs and UCS blade chassis, but I am no longer "on-loan" to the group running them and as such one hardware and probably several software revision out of date. Not to be awkward (any more than usual ) but I'd appreciate any expansion on the " If you're not using FI I'd look elsewhere" type comments. Thanks for the comments, some interesting feedback. Ones not on the list will probably work but might show errors in the management software. You can add after market RAM but they have a list of compatible chips (quite a few). Basically if you don't have 10gbit infrastructure you will wish you did because getting the fabric interconnects to talk 1gbit is a real pain.Īs for the servers themselves, they can be a bit finicky on RAM so you are best off getting what you want as part of the original purchase deal. With rack servers, that is a bit different but can still be the network choke point for the servers.
The fabric interconnects are basically kind of the brain for the server/chassis management and potentially for the networking as well. The interface for management is Java, which always pisses me off, but that is about the worst thing I can say about them. Of course, it depends on using things like IP allocation pools, MAC address pools, network storage, etc. The config is sort of based around templates so if you intend to have a good handful of servers that do the same chores, it is nice to be able to put in a new one and have it up and running really fast.