Diagnosing, and reporting latency issues

How to diagnose latency issues on your Crunchbits services.

What is Network Latency?

Network latency is the time it takes for data to travel from one point to another across a network. It's measured in milliseconds (ms) and represents the delay between sending a request and receiving a response.

Gathering the data

Ping, and MTR

When investigating latency issues, our network engineering team needs two different types of data:

Ping tests - these provide accurate latency measurements. They show the true round-trip time for packets traveling to and from the destination, giving us precise numbers on how much delay you're experiencing.

MTR - these reveal the complete network path and show where issues occur. While MTR's latency measurements can be skewed because routers often deprioritize ICMP responses used for traceroutes, it shows the route packets take, and helps us determine where the problem lies.

Installing MTR

Debian based OS

You can install MTR using the following commands:

sudo apt update
sudo apt install mtr

CentOS based OS

macOS

Windows

Running the tests

Please run MTR, and ping tests to the following IPs:

[The IP that you are seeing increased latency to]
[The gateway IP of your server]
1.1.1.1
8.8.8.8
104.36.84.66
66.23.198.57

Using the following command:

ping -c 200 [IP-from-the-list-above] > ping_results-[IP-from-the-list-above].txt
mtr -c 200 -r [IP-from-the-list-above] > mtr_results-[IP-from-the-list-above].txt

Bi-directional testing

Wherever possible, please make sure to run these tests both from your VPS to the destination, as well as from the destination, to your server. If the endpoint to which you are doing these tests isn't directly accessible by you, please check if they have a looking glass.

If the service that you are having latency issues with does have one, then please run a ping, and an MTR test from that looking glass to your server.

Alongside this, please also run MTR & ping tests from our looking glass to your server. This will help us determine if the issue comes from within our network, or externally.

What to do before opening a ticket

Please look at the results of your tests. If the latency is only increased to either a certain network, or only to a certain host, please reach out to the other side instead of us.

The next thing that you should do, is check our status page, and make sure that we currently don't have any open network outages, and or maintenance.

Another thing that you should also take a look at, is the status page of the other side. Occasionally, they might be under maintenance, which may increase your latency.

The last thing that you should do, is wait 24-48 hours, and redo the tests. Sometimes there might be an outage/degradation of the fiber line in between us, and the other side, which we are not always notified about. These are usually sorted out within 24-48 hours.

Opening a ticket

1. Open a ticket

Start by going to our client area, and then on the left navigation bar, please select Support [1] > Open Ticket [2]

3. Fill out the body of the ticket

Then, select the Technical Support department

Afterward, please fill out the Subject, and the Message part as per the table below:

Subject [1]

Latency issues - [xxx.xxx.xxx.xxx]

Message [2]

Hello,​ I'm currently encountering latency issues with my service IP: [xxx.xxx.xxx.xxx] These issues started on: [MM/DD/YYYY HH:MM PDT] The issues are happening to: [All endpoints, certain endpoints etc..] Below are the MTR & ping results: [From my service to XYZ Ping: ............ MTR: ............ From XYZ to my service: Ping: ............ MTR: ............] Please take a look at why this might be happening. Thank you!

In here please make sure to populate the variables within the square brackets with the appropriate values. You can also attach your tests as the .txt files that were automatically generated.

3. Submit the ticket

Click on the Submit button on the bottom of the page, and wait for us to take a look at your tests!