Home
» PC Tips
»
What is QoS? How to use QoS for faster Internet speeds when you need it most
What is QoS? How to use QoS for faster Internet speeds when you need it most
WebTech360 - In the article below, we will introduce and guide you to learn about the concept of QoS - Quality of Service on router devices.
What is QoS?
QoS (short for Quality of Service) is a way to control the traffic priority of the network system, this feature works on all different layers of the system, but in this test, we will focus on router devices used in home or personal models. More specifically, QoS will clearly demonstrate its effects in locations where bottlenecks often occur (or commonly called bottlenecks), and at the same time decide which part of the traffic is more important than the rest, based on the rules that the user sets up related to IP addresses, MAC addresses, active services...
QoS Use Cases in Practice
To understand it a little better, let's take the example of a traffic jam on a highway during rush hour. All the drivers on the road have one goal, to reach their final destination. And so they have to keep moving, even at a snail's pace.
Then the sound of an ambulance siren alerted them to a vehicle that needed to get to its destination urgently and had to go ahead of them. So the drivers moved out of the way - now the ambulance's "right-of-way" - and let it pass.
Similarly, when a network transmits data, it also has a setup where some types of data are treated better than all others. Important data packets need to reach their destination much faster than the rest, because they are time-sensitive and will “expire” if they don’t arrive on time.
Why is QoS important?
In the old days, information networks and business networks were separate entities. Telephone calls and remote communications were typically handled by an RJ11-connected network. Calls were monitored by a PABX system. This ran separately from the RJ45-connected IP network that connected laptops, desktops, and servers. The two types of networks rarely intersected, unless, for example, a computer needed a phone line to connect to the Internet. An example of such a network would look like this:
A computer needs a telephone line to connect to the Internet.
When networks were only transmitting data, speed was less important. Today, interactive applications that carry audio and video must be delivered over networks at high speeds, without packet loss or variation in delivery speed.
People now make business calls using online meeting applications like Skype, Zoom, and GoToMeeting, which use IP transport protocols to send and receive video or audio messages. For the sake of speed, these important applications can get the job done without the administrative processes that standard data transmissions typically require.
Where does bottleneck usually occur?
The main operating mechanism of QoS is usually applied when a bottleneck occurs at a certain time or place in the system, and the main factor here is the parameters you set related to bandwidth :
Suppose your QoS settings are set to exceed the bandwidth you are getting from your ISP. What happens if your router traffic is not prioritized because the system “thinks” that the bandwidth is reasonable. Meanwhile, if you have reached your ISP’s limit, they are the ones deciding what is and is not allowed to continue running.
Besides, if you set the QoS bandwidth level lower than the ISP's standard, it means you are creating an "artificial" bottleneck, and we can control, monitor and observe them through the router.
Some points to note to ensure stability in the system, including Uplink and Downlink settings :
Let's start by checking the average connection speed we are using from our ISP. To do this, you can use some available online support services such as Speakeasy Speed Test and SpeedTest :
According to the recommendations of experts as well as many experienced users, you should set QoS at about 85% of the value obtained from the test to ensure the system operates effectively. After determining the Uplink and Downlink speed , you should gradually increase it by about 1-2% each time. For some special cases, users can increase it up to 95% while the system still ensures performance, without being affected by any side effects.
What happens if QoS is not used in the network?
Not having QoS configured correctly can lead to one (or all) of the following problems:
Delay
When RTP packets are not assigned a priority, they are delivered at the default rate of the device. In a congested network, the packets have to travel along with the rest of the non-urgent packets. While network latency itself will not affect the quality of the audiovisual data being delivered, it will affect communication between end users. At 100ms of latency, one person will start talking over another as packets arrive out of sync, and at 300ms, the conversation will become incomprehensible.
Jitter
Real-time applications eliminate standard transport-level buffering, so there is no mechanism to reassemble incoming packets in the correct order. Jitter is the irregular rate at which packets arrive on the network. It can result in packets arriving late and out of sequence. Because the application does not wait for the stream to be properly ordered, out-of-sequence packets are dropped, resulting in distortion or gaps in the delivered audio or video.
Packet loss
Some (or part of) packets are lost due to too much congestion on network devices
This is the worst case scenario where some (or a portion) of packets are lost due to too much congestion on network devices. When the output queue of a switch or router is full, a tail drop occurs where the device drops any new packets until capacity is available again.
In all of the cases you just saw, QoS can help by scheduling data, managing queues, and preventing data loss.
It's not hard to imagine how communication and media streaming could be severely impacted without QoS - especially on networks serving the RTP protocol. Even if it's perfectly designed, eventually communication will first become difficult, then deteriorate as network traffic increases, and eventually become impossible.
The three errors mentioned above are actually very important in determining the quality of RTP-based traffic that QoS and network monitoring software companies like SolarWinds use as a metric.
It would be quite unfair to continue without mentioning a little more about one of the best network monitoring tools out there: SolarWinds NetFlow Traffic Analyzer.
This network monitoring suite helps solve problems that may be caused by:
- Slow Network : A slow network can impact a business as it continues to slow down data transfer speeds. Unless network bottlenecks are removed, the entire organization will experience poor connectivity.
- Slow communication : A business that fails to establish a clear communication channel within its network will be paralyzed. Worse still, failing to communicate clearly with customers will almost certainly get the business into trouble.
- Unmonitored Network : Administrators who cannot properly monitor their networks will not be able to know the current state of their networks or how to plan for future network expansion. Without monitoring the network and tracking the performance of each device, network managers cannot make informed decisions and are likely to exacerbate network performance issues.
Equipped with Netflow Traffic Analyzer, network administrators will be able to solve commonly encountered problems, as the tool helps:
Helps implement QoS and optimize it - through data flow feedback
Review and report on current QoS policy configuration, informing design decisions.
Monitor bandwidth usage to determine which applications and devices are taking up network resources - these can be isolated, rescheduled, or disabled.
A typical Netflow Traffic Analyzer dashboard contains important information that administrators need to monitor statuses and make quick settings adjustments. An example:
SolarWinds NetFlow Traffic Analyzer tool example
These reports and analytics include: Latency, Jitter, and Packet Loss.
Another option you can use to monitor QoS is Paessler PRTG. This network monitoring toolkit has a dedicated section for monitoring QoS performance. This feature shows you tagged traffic flows in real time, and it also stores the data for performance analysis and capacity planning.
PRTG software includes 4 monitoring sensors, covering 3 different QoS methods. They are supplemented with a Ping Jitter sensor that monitors the regularity of packet delivery within a flow.
The three types of QoS that PRTG can monitor are Standard QoS, Cisco IP-SLA, and Cisco CBQoS. Standard QoS monitors are implemented as one-way sensors or Round Trip sensors. These monitors can operate on connections across the Internet.
Paessler PRTG
Cisco IP-SLA Sensor is specifically designed to monitor tagged VoIP traffic on your network. It records a variety of metrics for voice traffic, including round-trip delay, latency, jitter, and Mean Opinion Score (MOS).
Cisco CBQoS Sensor follows a class-based Quality of Service implementation. CBQoS is a queuing method and if you want to implement it, you will have to monitor multiple entry points on your routers and switches. You create at least 3 virtual queues for each device, so there is more to monitor.
PRTG can automatically set up and map all of your network infrastructure. However, implementing QoS requires decision making, so you will have to set up the method yourself by deciding which types of network traffic to prioritize.
Paessler lets you use PRTG for free if you only activate up to 100 sensors. If you need more, you can try the system for free for 30 days, which includes QoS monitoring.
How to configure QoS?
Cable routers and switches can be configured to prioritize protocols that are typically accessed by the router management software suite. The entire process of configuring QoS options is a fairly straightforward task that includes:
Log into the app and connect to the hub or switch through it
Navigate to the QoS configuration menu
Set package priority options
And thus, media packets will be able to flow smoothly over the network. Hardcore network engineers can perform all the tasks listed above through the command line configuration interface.
How are RTP packets prioritized?
QoS packet prioritization can be done using two main methods:
- Classification : This efficient method identifies packet types and assigns their priority by marking them. The identification can be done using ACLs (Access Control Lists), implementing LANs using CoS (Class of Service) or with the help of switches using hardware-based QoS marking.
- Queues : Queues are high-performance memory buffers found in routers and switches. Packets passing through them are held in dedicated memory areas as they wait to be sent. When protocols, such as RTP, are assigned a higher priority, they are moved to a dedicated queue that pushes data forward at a faster rate, thus reducing the risk of being dropped. Lower priority queues do not get this treatment.
One important thing to remember here is that the priority marking of a packet is only valid within the network it was created in. Once it leaves the network, the owner of the receiving network determines its new priority.
Things to consider when prioritizing packets
Some things and tips that can be helpful when deciding how to prioritize packages include:
- It is generally a good idea to have priority markers assigned by the devices closest to the data source. This ensures that packets travel across the entire network with the correct priority.
- The device chosen to mark incoming packets should always be a switch. This is because these devices can load balance traffic on the network and share the burden with other switches, thus reducing the burden on their CPUs.
- Incoming traffic is almost always larger than outgoing traffic. ISPs usually allocate less bandwidth to customer outgoing traffic and it is there (on the outgoing network path) that QoS mainly needs to be applied.
Cisco has a recommendation for how packets should be marked as shown in this diagram:
Cisco's Recommendations on How Packets Should Be Marked
Ultimately, the success of a QoS implementation always depends on the quality of the policy that governs how packets are classified, marked, and queued. The policy must be carefully crafted for a successful QoS implementation.
What should QoS not be used for?
After reading about QoS, you might think of it as a panacea for network congestion. To a certain extent, QoS can make most RTP communications smoother and seemingly streamline network traffic. Unfortunately, QoS is not a one-size-fits-all solution to every network problem.
Never use QoS for the following purposes:
Increase bandwidth
While QoS helps prioritize RTP packets and makes it look like the network is suddenly increasing bandwidth, it should never be taken that way. QoS should never be used as a tool to “increase bandwidth,” when all it does is use existing resources a little more efficiently (and in favor of RTP packets).
Instead, consider caching files to reduce the amount of data coming and going. If that doesn't work, it means that the bandwidth limit has been reached. When a company reaches its bandwidth limit, the only viable option is to go out and buy some more network equipment, not use QoS.
Keep the network from getting congested
If some applications are still running and are slowly consuming network bandwidth, implementing QoS is not the solution. While Skype calls may finally start to get through, QoS will not address the root problem. Ultimately, the applications will “swallow” whatever resources are available, negating the benefits of QoS.
One solution that might work here is to look for resource-hungry apps and disable them or reschedule them to run after hours.
Again, the whole purpose of configuring QoS on your network is to ensure that your streaming audio and video calls don’t lag (or even crash) due to network congestion. QoS isn’t a tool that can actually increase bandwidth. It also can’t get through a congested network.
A good QoS implementation improves the quality and speed of critical data by optimizing the allocated bandwidth, facilitating the tagging of packets so they are identified, and assigning specified priorities. QoS uses the available bandwidth, but does not expand it.
Frequently Asked Questions about QoS in Networking
What is the difference between QoS and Network Throttling?
Throttling, also known as policing, involves setting an overall limit on traffic throughput and reducing excess traffic levels. QoS is a method of prioritizing some traffic over others and using queuing, thereby maximizing bandwidth for some traffic at the expense of bandwidth for other traffic.
What is the main role of DSCP in QoS?
Differentiated Services Code Point (DSCP) appears in the packet header. This is a packet-level opportunity to request priority from the QoS management software on the network devices. The network manager can choose to enable or disable DSCP discovery on the device, so this value can be ignored in favor of another QoS queuing method.
Can you explain traffic shaping in QoS?
Traffic shaping is a method used by QoS to get the best value from network capacity. All networks experience peak demand and traditional capacity planning requires provisioning bandwidth at the peak level plus a safety margin. QoS traffic shaping introduces small delays on certain traffic to allow a network with less capacity than peak demand to accommodate all traffic.