I’m pleased to revive the KIC LABS series with a video on Cisco MQC-style QoS on the Nexus 5000.
PLEASE read these two articles first, as they do a lot better job at explaining the structure that we’re going to be putting together in the video. Unfortunately I was not able to provide this during the video, as I was pressed for time.
https://keepingitclassless.net/2012/11/cisco-quality-of-service-part-1-types-of-qos-policies/
https://keepingitclassless.net/2012/11/qos-part-2-qos-and-jumbo-frames-on-nexus-ucs-and-vmware/
Download the configuration used in this video here.
I’m pleased to revive the KIC LABS series with a video on Cisco MQC-style QoS on the Nexus 5000.
PLEASE read these two articles first, as they do a lot better job at explaining the structure that we’re going to be putting together in the video. Unfortunately I was not able to provide this during the video, as I was pressed for time.
https://keepingitclassless.net/2012/11/cisco-quality-of-service-part-1-types-of-qos-policies/
https://keepingitclassless.net/2012/11/qos-part-2-qos-and-jumbo-frames-on-nexus-ucs-and-vmware/
Download the configuration used in this video here.
This started off as a company email but I wanted to share it, since I’ve been asked before. Below is opinion and opinion only. I’m more interested in how this compares with your study methods and Cisco exam experiences, so please let me know your thoughts in the comments.
Cisco exams….let’s face it, they suck. (In a good way though) I’ve taken exams from all kinds of vendors, and Cisco is really good at creating exams that vet you technically, but also require that you understand the practical reasons for the technologies we work with, not just memorizing maximum values, or configuration steps.
I’ve run into many circumstances where jumbo frames are enabled, most notably in the data center. After all, allowing for a maximum tranmission unit of greater than 1500 bytes allows us to transmit more data per frame
As I explained in Part 2 of my QoS Series, MTU can be a touch subject. Do it wrong, and you encounter one of two big network problems. One potential issue when configuring jumbo frames at L2 is that stuff just doesn’t work.
Yesterday kicked off the 5th iteration of Network Field Day. For those that haven’t heard of Tech Field Day, you need to check it out - there’s something for everyone, and it’s a great event that gets the technical details from vendors on their solutions. The delegates that are invited are what I consider thought leaders in each field. I’ve had the privileged of blogging, podcasting, and even meeting with them in person over the past few years, and they’re just the right kind of folks to help keep these vendors honest.
I’ve had some great conversations lately with a lot of folks on the topic of IPv6 prefix length in a variety of applications, specifically one very good discussion on just about anything IPv6 between me, the kind folks over at The Class-C Block and Tom Hollingworth (aka The Networking Nerd).
For many folks that are considering the impact of going dual-stack in their environments, the idea of using a /64 on all links is still a point of contention.
Now we arrive at the question of how much address space to allocate for…anyone. You may be a service provider, you may be a business, you may be a home user. Today, this question is quite easy to solve. If you’re a business-class customer, you ask your ISP for a block of addresses, and based off of your need (or ability to justify the need), you’ll be allocated some addresses. For many small-to-medium businesses, this can be as small as 8, or even 4 addresses.
As a Data Center junkie, I daily bear witness to the glorious transformations that are taking place all around me with respect to the “next-generation” of data center. Everyone who wants to move their DC to the next level are millions of dollars worth of DC networking gear that is EXTREMELY cutting edge, enabling virtualization and cloud to do things we only dreamed of being able to do mere years ago.
For so long, I’ve heard - as have many of you I’m sure - that GNS3, though a GREAT emulator for Cisco IOS software, is not practical for studying anything related to switching. Routing is handled just fine, but because of the proprietary ASICs in Cisco switches, it is not something that can be easily reverse-engineered, thus GNS3 cannot do it. After all, all routing is essentially done in software in GNS3.
I have seen the conversation around SDN evolve over what amounts to the last few years from something that was barely whiteboard material, to something on everyone’s lips in this industry. Why? What’s so interesting about these three little letters? Well, if you’ve heard of it, you’ve undoubtedly heard from your local vendor account manager that their product is the leader in the SDN market, or that they just made a big acquisition that really puts them ahead in the SDN space, blah, blah, blah.
The idea of security in a vSphere vSwitch is a concept not usually discussed in vSphere peer groups or curricula. They are somewhat specialized features that are normally either not used, or irrelevant due to the presence of another switching architecture such as the vDS (including the Cisco Nexus 1000v) or VM-FEX, where these policies also exist and are much more feature-rich. Thus, the idea of performing these functions on a native vSwitch is usually not talked about.
I’m working on setting up a lab that consists of leading storage and compute products for testing, and I ran into some interesting issues with a few different things…some with respect to the way the Cisco ASA does hairpinning, as well as allowed connections in such a configuration. There were also some routing issues experienced as a result, and I want to explore my experience in all of this during this post.
I’m working on setting up a lab that consists of leading storage and compute products for testing, and I ran into some interesting issues with a few different things…some with respect to the way the Cisco ASA does hairpinning, as well as allowed connections in such a configuration. There were also some routing issues experienced as a result, and I want to explore my experience in all of this during this post.
Last week I did an overview of the performance-minded storage solution that Nexgen has put together. In summary, by using SSD-based read AND write caching that’s moved in and out of the cache in an intelligent way, we can get better performance than traditional disk arrays with slower disks, and fewer of them. I’d like to do a quick tour of the hardware for their low-end model, the n5-50. It’s actually pretty straightforward and the internals are interesting enough that I decided to take some pictures and discuss their role in the solution.
Last week I did an overview of the performance-minded storage solution that Nexgen has put together. In summary, by using SSD-based read AND write caching that’s moved in and out of the cache in an intelligent way, we can get better performance than traditional disk arrays with slower disks, and fewer of them. I’d like to do a quick tour of the hardware for their low-end model, the n5-50. It’s actually pretty straightforward and the internals are interesting enough that I decided to take some pictures and discuss their role in the solution.
I began my professional career after college by starting in route/switch. Although I still do plenty of route/switch work, I have also recently taken on responsibilities focused on more datacenter-centric technologies like virtualization, and storage, in addition to the networking in the back-end that makes it all work. Much to the stress of my schedule (and my……stress), one has not trumped the other - they simply exist in parallel.
I was given the privilege to tinker with some gear from my friends over at Nexgen Storage. For those that have not heard of them, I encourage you to head over to http://www.nexgenstorage.com/product/technology and take a peek at the solution. They are one of the “little guys”, but they’re doing some cool things with respect to performance, and providing the ability to give priority to certain tiers of applications or tenants that are using the system.
I was given the privilege to tinker with some gear from my friends over at Nexgen Storage. For those that have not heard of them, I encourage you to head over to http://www.nexgenstorage.com/product/technology and take a peek at the solution. They are one of the “little guys”, but they’re doing some cool things with respect to performance, and providing the ability to give priority to certain tiers of applications or tenants that are using the system.
I set up Auto Deploy in my home lab using vSphere 5.1 on an existing server, in order to boot a Cisco UCS C220 M3 server whose local hard drives have not arrived yet.
I followed Duncan Epping’s walkthrough for Auto Deploy on vSphere 5.0, but this post is about what I had to do differently to get it working. Hopefully I save you some headaches. There might be some improvements to this process, but I was under a deadline and I know that it worked for me - please share any improvements in the comments.
2012 has been a crazy year for me. I’d like to briefly summarize my year and publicly post some of my goals for 2013.
Accomplishments in 2012 Blogging - Monthly Views to Keeping It Classless increased by over 450% this year, and every single month had consistently more views than the month before. I want to thank each and every one of you for reading my articles - I really only got into this recently and the explosive growth is still hard to believe.
Keeping It Classless