Jess Probasco: upgrading core switches no longer requires network downtime

Jess Probasco: upgrading core switches no longer requires network downtime

Network downtime often occurs during network upgrades, especially in the 7x24 broadcast media environment, which has a significant impact.

No one prefers to stop, but this is inevitable during network upgrades. For most network engineers, the appearance of downtime windows at night may have little impact on users. However, in the broadcast media, these downtime windows are hard to handle. To this end, we talked to Jess Probasco, a senior network engineer at American Public Media. He shared his experiences and feelings about network maintenance at the second largest Public broadcasting manufacturer in the United States.

What have you been busy with recently?

Jess Probasco:We recently used IP multicast in the studio to debug audio for our broadcast and recording, and just changed a large core switch and all access switches for this environment. We completed the encapsulation last month and now we are configuring the Infrastructure Section. We have just upgraded to the Cisco Catalyst 3850 switch and the Cisco Nexus 7000 core switch.

Why do we invest in Cisco's vswitches?

Jess Probasco:We have configured a Cisco Catalyst 6500 series switch. It has been in use for 8 years and will be unavailable immediately. All of our edge switches have been in use for almost years, so we need to find an environment that can provide us with high availability and redundancy, in addition, we can use some new technologies that can make full use of this environment. Now, we are preparing to add virtual servers and Cisco UCS racks for broadcast devices and servers in our broadcast environment. Previously, we deployed Nexus 7000 in the data center, which ensures that the technology is the same in both environments.

How does a broadcast media company maintain the network?

Jess Probasco:After the upgrade, you do not need to stop the service in many cases. When faced with audio or online broadcast, as we expected, we can maintain the downtime window and planned downtime, making it difficult to find other broadcast interruptions. For example, when upgrading, we can complete the entire core switch, edge switch, and all other upgrades within five seconds. The downtime becomes short, which is why we are forced to configure a new Nexus environment with dual-switch, dual-core, and high availability.

Every day, there is a constant jump from one ISP to another due to weather interruptions. There are also some problems that our wireless launch sites are often not fixed, which is also a part of the challenge. If we can connect to some sites stably, especially after the storm, it would be better.

Data is usually normal, but it is a typical network problem. However, audios are all multicast-driven, so sometimes problems are difficult to solve. The problems may not be a single point.

What new technologies are being learned recently?

Jess Probasco:I recently studied software-defined networks, which are more programming than traditional exchange technologies. Combined with some automation tools, you can use them to display, configure, and change vswitches. Therefore, it is also good to know some non-SDN platform scripting languages. I know that Cisco is doing better in Python scripts and can be used in Cisco Prime and Cisco ACS products. So now I am most interested in learning some programming languages. Wherever you go, this is a trend-network programmability.

Can you give an example of programming in today's networks?

Jess Probasco:We are currently in the process of Cisco Prime infrastructure launch, so we need to be programmable in large networks using new switches, or you can share some of the daily work that can help the desktop or the underlying layer. We hope to say, 'Hi, someone on DataSet A or dataset B wants to access the VLAN or needs to challenge the port to another VLAN. 'In this way, we can provide them with this capability or do not need to manually jump to the vswitch. This is what we call programmability.

Have you ever thought about what you want to do when you grow up? How did I become a network engineer?

Jess Probasco:Like many children, I had many dreams when I was a child. I first wanted to work as a medical engineer, and later I wanted to work as an electrical engineer. When I was 15 years old, I had a chance to go to an electronic engineering company for an internship, but later I found that it was not my path. I didn't want to play with cables and circuits all day. So I started to study computers and started on this path. I first came into contact with the voice. Later, with the development of VoIP, I switched from the voice to the network. Since then, I became fascinated by the network and was very interested in what the network could do.

If you are a superhero, what capabilities do you want to possess?

Jess Probasco:This is hard to say. But if I am a superhero, I hope I can be in multiple places at the same time.

It sounds useful at work.

Jess Probasco:Yes, but it's not just work. Of course, it's also used in life. Sometimes we are too busy, so it would be nice if we could stay in multiple places at the same time.

[Edit recommendations]

Techniques for managing and maintaining vrouters in line management mode maintained by individual modules in front of cabinets SDN market: opportunities or crises for channel partners? How enterprises balance IT maintenance and innovation and talk about the daily maintenance work of data centers [responsible editor: Lin lintel :( 010) 68476606]