You'll learn how the Juniper QFX enables you to create simple-to-use data centers or build some of the largest IP Fabrics in the world. This book is. Like the popular guides The MX Series and Juniper QFX Series, this practical book—written by the same author—introduces new QFX concepts in. GitHub Repo for the O'Reilly QFX book by Doug Hanks - Juniper/qfx book.
|Language:||English, French, Portuguese|
|ePub File Size:||26.52 MB|
|PDF File Size:||13.67 MB|
|Distribution:||Free* [*Registration needed]|
You'll learn how the Juniper QFX Series enables you to create simple-to- use data centers or build some of the largest IP Fabrics in the world. pages. Juniper Networks books on routing and certification address the unique requirements of service provider and carrier customers. With this book, you'll be well on your way to becoming a Juniper .. The Juniper QFX series of switches is quickly becoming the go-to platform for a.
With its wide variety of features, the Juniper QFX is able to quickly solve the challenges of cloud computing as well as other use cases such as high-frequency trading and campus. Junos is a purpose-built networking operating system based on one of the most stable and secure operating systems in the world: Junos is designed as a monolithic kernel architecture that places all of the operating system services in the kernel space.
Major components of Junos are written as daemons that provide complete process and memory separation.
One of the benefits of monolithic kernel architecture is that kernel functions are executed in supervisor mode on the CPU, whereas the applications and daemons are executed in user space. A single failing daemon will not crash the operating system or impact other unrelated daemons. Creating a single network operating system that you can use across routers, switches, and firewalls simplifies network operations, administration, and maintenance.
Network operators need only learn Junos once and become instantly effective across other Juniper products. Being able to write these core protocols once and then reuse them across all products provides a high level of stability because the code is very mature and field tested. Every quarter for more than 15 years, there has been a consistent and predictable release of Junos.
The development of the core operating system is a single-release train. This allows developers to create new features or fix bugs once and then share them across multiple platforms. The release numbers are in a major and minor format. The major number is the version of Junos for a particular calendar year, and the minor release indicates in which trimester the software was released. There are a couple of different types of Junos that are released more frequently to resolve issues: Service releases are released on demand to specifically fix a critical issue that has yet to be addressed by a maintenance release.
The general rule of thumb is that new features are added for every major and minor releases and bug fixes are added to service and maintenance releases. For example, Junos Most production networks prefer to use the last Junos release of the previous calendar year; these Junos releases are Extended End of Life EEOL releases that are supported for three years.
The advantage is that the EEOL releases become more stable with time. Consider that This increased the frequency of maintenance releases to resolve more issues more often. The other benefit is that all Junos releases as of are supported for 24 months, whereas the last release of Junos for the calendar year will still be considered EEOL and have support for 36 months. By extending the engineering support and reducing the number of releases, network operators should be able to reduce the frequency of having to upgrade to a new release of code.
With the new Junos three-release cadence, network operators can be more confident using any version of Junos without feeling pressured to only use the EEOL release. Junos was designed from the beginning to support a separation of control and forwarding plane. For example, the forwarding plane could be forwarding traffic at line rate and performing many different services while the routing engine sits idle and unaffected.
Control plane functions come in many shapes and sizes. In fact, there are many more control plane functions. Some examples include:. At a high level, the control plane is implemented entirely within the routing engine, whereas the forwarding plane is implemented within each PFE using a small, purpose-built kernel that contains only the required functions to forward traffic.
The benefit of control and forwarding separation is that any traffic that is being forwarded through the switch will always be processed at line rate on the PFEs and switch fabric; for example, if a switch were processing traffic between web servers and the Internet, all of the processing would be performed by the forwarding plane.
The Junos kernel has five major daemons. Each of these daemons play a critical role within the Juniper QFX and work together via Interprocess Communication IPC and routing sockets to communicate with the Junos kernel and other daemons. These daemons, which take center stage and are required for the operation of Junos, are listed here:. The sections that follow provide descriptions of each of the five major daemons. This makes it possible for Junos to handle data in interesting ways and opens the door to advanced features such as configuration rollback, apply groups, and activating and deactivating entire portions of the configuration.
The UI has four major components: The management daemon is the glue that holds the entire Junos UI together. At a high level, it provides a mechanism to process information for both network operators and daemons. The interactive component of the management daemon is the Junos CLI. This is a terminal-based application that provides the network operator with an interface into Junos. This provides an API through Junoscript and Netconf to accommodate the development of automation applications.
To see an example of this, simply add the pipe command display xml to any command. The routing protocol daemon handles all of the routing protocols configured within Junos. At a high level, its responsibilities are receiving routing advertisements and updates, maintaining the routing table, and installing active routes into the forwarding table. To maintain process separation, each routing protocol configured on the system runs as a separate task within the routing protocol daemon. Its other responsibility is to exchange information with the Junos kernel to receive interface modifications, send route information, and send interface changes.
The hidden command set task accounting toggles CPU accounting on and off. Use show task accounting to see the results:. Currently, running daemons and tasks within the routing protocol daemon are present and accounted for.
The set task accounting command is hidden for a reason. The device control daemon is responsible for setting up interfaces based on the current configuration and available hardware. One feature of Junos is the ability to configure nonexistent hardware.
However, as soon as hardware is installed into FPC1, the first port will be configured immediately with the address The chassis daemon supports all chassis, alarm, and environmental processes. At a high level, this includes monitoring the health of hardware, managing a real-time database of hardware inventory, and coordinating with the alarm daemon and the craft daemon to manage alarms and LEDs.
It should all seem self-explanatory except for the craft daemon, the craft interface that is the front panel of the device. This information can also be obtained via the command line as well with the command show chassis led , as illustrated here:. One final responsibility of the chassis daemon is monitoring the power and cooling environmentals. It constantly monitors the voltages of all components within the chassis and will send alerts if any of those voltages cross specified thresholds.
The same is true for the cooling. The chassis daemon constantly monitors the temperature on all of the components and chips as well as fan speeds. If anything is out of the ordinary, it will create alerts. Under extreme temperature conditions, the chassis daemon can also shut down components to avoid damage. To help answer these questions, the Juniper QFX brings a new daemon into the mix: The data collected can be broken down into two types:.
Each port on the switch has the ability to queue data before it is transmitted. The ability to queue data not only ensures the delivery of traffic, but it also impacts the end-to-end latency. The analytics daemon reports data on the queue latency and queue depth at a configured time interval on a per-interface basis. Being able to measure the packets per second pps , packets dropped, port utilization, and number of errors on a per-interface basis gives you the ability to quickly graph the network.
You can access the data collected by the analytics daemon in several different ways. You can store it on the local device or stream it to a remote server in several different formats. Traffic must be collected from two locations within the switch in order for the data to be accurate. The second location to collect traffic is from the routing engine. The PFE sends data to the routing engine if that data exceeds certain thresholds.
The analytics daemon will then aggregate the data. The precise statistics directly from the PFE and the aggregated data from the routing engine is combined to give you a complete, end-to-end view of the queue and traffic statistics of the network. Routing sockets are a UNIX mechanism for controlling the routing table. The Junos kernel takes this same mechanism and extends it to include additional information to support additional attributes to create a carrier-class network operating system.
At a high level, there are two actors that use routing sockets: The routing protocol daemon is responsible for processing routing updates and thus is the state producer.
Using the rtsockmon command from the shell allows us to see the commands being pushed to the kernel from the Junos daemons. The command rtsockmon is a Junos shell command that gives the user visibility into the messages being passed by the routing socket.
The routing sockets are broken into four major components: The sender field is used to identify which daemon is writing into the routing socket. The type identifies which attribute is being modified. The operation is showing what is actually being performed. There are three basic operations: The last field is the arguments passed to the Junos kernel.
The rtsockmon command is used only to demonstrate the functionality of routing sockets and how daemons such as dcd and rpd use routing sockets to communicate routing changes to the Junos kernel.
Each has varying numbers of ports and modules, but they share all of the same architecture and benefits. Depending on the number of ports, modules, and use case, a particular model can fit into multiple roles of a data center or campus architecture. In a spine-and-leaf architecture, this model is most commonly deployed as a spine fulfilling the core and aggregation roles. There are no modules, but there is enough bandwidth to provide 2: There is Gbps of downstream bandwidth from the 48 10GbE interfaces and Gbps of upstream bandwidth from the 6 40bGE interfaces.
In a spine-and-leaf architecture, this model is most commonly deployed as a leaf fulfilling the access role. In a spine-and-lead architecture, this model is most commonly deployed as a leaf.
Depending on how many modules and which specific module is used, the port count can change for the models that have expansion ports. For example, the Juniper QFXQ has 24 40GbE built-in interfaces, but using two modules can increase the total count to 32 40GbE interfaces, with the assumption that each module has 4 40GbE interfaces.
Each model has been specifically designed to operate in a particular role in a data center or campus architecture but offer enough flexibility that a single model can operate in multiple roles. The modules make it possible for you to customize the Juniper QFX series to suit the needs of the data center or campus.
Depending on the port count and speed of the module, each model can easily be moved between roles in a data center architecture. Using this module, you can add an additional Gbps of bandwidth via 4 40GbE interfaces. You can use the interfaces as-is or they can be broken out into 16 10GbE interfaces with a breakout cable.
This module adds an additional 80 Gbps of bandwidth via 8 10GbE interfaces. You typically use the 8 10GbE module to add additional downstream bandwidth for connecting compute resources. In a data center architecture, it can fulfill the roles of the core, aggregation, and access. The reason the Juniper QFXQ is able to collapse the core and aggregation roles is because it offers both high-speed and high-density ports in a single switch.
The combination of built-in ports and modules brings the total interface count to 32 40GbE. These switches are providing 40GbE access interfaces to Host 1. The Juniper QFXQ is a very flexible switch that you can deploy in a variety of roles in a network.
Although the math says that with 32 40GbE interfaces you should be able to get 10GbE interfaces, the PFE has a limitation of total interfaces at any given time. This way, both the fans and power supplies have the same airflow, and the switch is cooled properly. Mismatching the airflow could result in the switch overheating.
The switch is powered by two power supplies. A really great feature of the Juniper QFX is the colored plastic on the rear of the switch. The handles to remove the fans and power supplies are color-coded to indicate the direction of airflow.
Blue represents cool air coming into the rear of the switch, which creates a back-to-front airflow through the chassis.
Orange represents hot air exiting the rear of the switch, which creates a front-to-back airflow through the chassis. Being able to quickly identify the type of airflow prevents installation errors and gives you peace of mind.
These components are management, cooling, and power. A power supply can experience a failure, and the other power supply has enough output to allow the switch to operate normally.
If the alarm is red, this is an indication that one or more hardware components have failed or have exceeded temperature thresholds. An amber alarm indicates a noncritical issue, but if left unchecked, it could result in a service interruption.
This LED is always green but has three illumination states: It is here to help remote hands and the installation of the switch; you can use it to help identify a particular switch with a visual indicator. If the ID LED is off, this is the default state and indicates that the beacon feature is currently disabled on the switch.
There are three management ports in total, but you can use only two at any given time; these are referred to as C0 and C1. Basically, the two C0 management ports are interchangeable, but you can use only one at any given time. The two management ports C0 and C1 are used for out-of-band management. In a QFabric architecture, each node and interconnect requires two out-of-band management connections to ensure redundancy. Having both a SFP and copper management port gives you more installation flexibility in the data center.
If you prefer fiber, you can easily use just the C1 interface and leave C0 unused. The RS console port is a standard RJ interface. This serial port is used to communicate directly with the routing engine of the switch. For situations in which the switch becomes unreachable by IP, the serial RS is always a nice backup to have. In a data center architecture, it has been designed to fulfill the role of the access tier. The primary role for the Juniper QFXS is to operate in the access tier of a data center architecture, due to the high density of 10GbE ports.
The 48 10GbE interfaces are generally used for end hosts, and the 6 40GbE interfaces are used to connect to the core and aggregation.
Juniper QFX5100 Series Book 1st Edition
In data centers where the end hosts are only 1GbE, you can change the roles of the Juniper QFXS and use it as a spine switch in the core and aggregation tiers of a data center architecture. The same logic holds true for a 1GbE spine-and-leaf topology: The key to a spine-and-leaf network is that the upstream bandwidth needs to be faster than the downstream bandwidth to ensure an appropriate level of over-subscription.
The primary role for the Juniper QFXT is to operate in the access tier of a data center architecture, due to the high density of 10GbE ports.
The Juniper QFXT is a very flexible switch in the access layer; network operators can use the same switch for both management and production traffic.
Typically, management traffic is Mbps or 1 Gbps over copper by using the RJ interface. An annual anal Embed Size px. Start on. Show related SlideShares at end.
WordPress Shortcode. Published in: Full Name Comment goes here. Are you sure you want to Yes No. Be the first to like this.
Views Total views. Actions Shares. Embeds 0 No embeds. No notes for slide.
Juniper QFX5100 Series: A Comprehensive Guide to Building Next-Generation Networks
There are five preconfigured profiles that range between L2 heavy to L3 heavy. Network Analytics Some applications are sensitive to microbursts and latency. The Juniper QFX allows you to get on-box reporting of queue depth, queue latency, and microburst detection to facilitate and speed up the troubleshooting process.
Virtual Chassis Fabric Ethernet fabrics provide the benefit of a single point of management, lossless storage convergence, and full Layer 2 and Layer 3 services. The Juniper QFX family brings a lot of new features and differentiation to the table when it comes to solving data center challenges. Because of the wide variety of features and differentiation, you can integrate the Juniper QFX into many different types of architectures.
High-Frequency Trading Speed is king when it comes to trading stocks.
Juniper QFX5100 Series
With an average port-to-port latency of nanoseconds, the Juniper QFX fits well in a high-frequency trading architecture. Private Cloud Although the Juniper QFX was specifically designed to solve the challenges of cloud computing and public clouds, you can take advantage of the same features to solve the needs of the private cloud. Enterprises, government agencies, and research institutes are building out their own private clouds, and the Juniper QFX meets and exceeds all their requirements.
Campus High port density and a single point of management make the Juniper QFX a perfect fit in a campus architecture, specifically in the core and aggregation roles.
Enterprise Offering the flexibility to be used in multiple deployment scenarios, the Juniper QFX gives an enterprise the freedom to use the technology that best fits its needs.
With its wide variety of features, the Juniper QFX is able to quickly solve the challenges of cloud computing as well as other use cases such as high-frequency trading and campus. Junos Junos is a purpose-built networking operating system based on one of the most stable and secure operating systems in the world: FreeBSD.
Junos is designed as a monolithic kernel architecture that places all of the operating system services in the kernel space. Major components of Junos are written as daemons that provide complete process and memory separation. One of the benefits of monolithic kernel architecture is that kernel functions are executed in supervisor mode on the CPU, whereas the applications and daemons are executed in user space.
A single failing daemon will not crash the operating system or impact other unrelated daemons. One Junos Creating a single network operating system that you can use across routers, switches, and firewalls simplifies network operations, administration, and maintenance.
Network operators need only learn Junos once and become instantly effective across other Juniper products. Being able to write these core protocols once and then reuse them across all products provides a high level of stability because the code is very mature and field tested.
Software Releases Every quarter for more than 15 years, there has been a consistent and predictable release of Junos.
The development of the core operating system is a single-release train. This allows developers to create new features or fix bugs once and then share them across multiple platforms. The release numbers are in a major and minor format. The major number is the version of Junos for a particular calendar year, and the minor release indicates in which trimester the software was released.
There are a couple of different types of Junos that are released more frequently to resolve issues: maintenance and service releases.The Junos kernel has five major daemons. At a high level, it provides a mechanism to process information for both network operators and daemons. A single failing daemon will not crash the operating system or impact other unrelated daemons. All of the examples and features are based on Junos releases Cooling The Juniper QFX family was designed specifically for the data center environment; each system supports front-to-back cooling with reversible airflow.