NEW

BLOG

SLA measurement high importance
Allprobe Technical Writer
2019-03-19 18:39:42

SLA, what is it & why is it critical to regularly sample it.


SLA came from the initial term "service-level agreement" presenting a company commitment for x% service availability, but you can refer to it as quality, availability, response sensitivity checks from an IT perspective. From a monitoring point of view, SLA should give you a precise number as a percentage out of 100% when 100% is the maximum value, and 0% means 0% availability of the sampled object usually a Host(with network connectivity).


A simple example of SLA in practice monitoring is PING check. Ping is a network availability tool used to test the reachability of a host. When regularly sampling a host, we can save and calculate the availability of those sampled host and have a good sense of this host availability quality.


Example for ping output:


PING google.com (216.58.206.14) 56(84) bytes of data.
64 bytes from 1e120.net (216.58.206.14): icmp_seq=1 ttl=53 time=76.0 ms
64 bytes from 1e120.net (216.58.206.14): icmp_seq=2 ttl=53 time=77.2 ms
64 bytes from 1e120net (216.58.206.14): icmp_seq=3 ttl=53 time=75.0 ms
64 bytes from 1e120.net (216.58.206.14): icmp_seq=4 ttl=53 time=80.4 ms
64 bytes from 1e120.net (216.58.206.14): icmp_seq=5 ttl=53 time=97.1 ms


--- www.example.com ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4004ms
rtt min/avg/max/mdev = 75.027/81.187/97.183/8.211 ms


It's hard to imagine the world without the internet these days but not much of us really familiar with the topology of the internet, the internet is made of many networking junctions, huge data centers, millions of cables and optic fibers going on the ground and under oceans and seas, satellites and even more bizarre ways to deliver cheap or even free connectivity to far places.


When measuring the availability rate of essential and international services relying on servers, cloud servers, virtual machines and many other types of hosts with some purpose, its also very beneficial to check it from various geo locations and network junctions to be able to see the full picture and spot issues with the networking status of the monitored host. Ultimately, those checks and gathered statistics will give a broader picture of your overall service from a networking point of view.


SLA can also be implemented for other types of checks other then PING and network availiilty, those checks are often application PORTs and HTTP calls.


SLA is also ubiquitous and essential for various API health checks, so, for example, one use case can include checking the response time of some API calls such as:


https://api.yourdomain.com/webservice/clients/call1
https://api.yourdomain.com/webservice/clients/call2
https://api.yourdomain.com/webservice/clients/call3


Those URLs can be a web service, microservice, API, a website, blogs, a service, and other types of data delivery point.


Tracking the fluctuations of various simple and not invasive checks such as HTTP, PORT, PING, are the foundation and a significant matrix to consider when measuring IT services & microservices environments.


An example of SLA report taken from my Allprobe BETA account

Sample SLA report



Our October 2019 significant UI and UX update
Discovery flowchart
CMDB & IT transformation
SLA measurement high importance
CMDB made easy
SLA measurement high importance
Allprobe Technical Writer
2019-03-19 18:39:42

SLA, what is it & why is it critical to regularly sample it.


SLA came from the initial term "service-level agreement" presenting a company commitment for x% service availability, but you can refer to it as quality, availability, response sensitivity checks from an IT perspective. From a monitoring point of view, SLA should give you a precise number as a percentage out of 100% when 100% is the maximum value, and 0% means 0% availability of the sampled object usually a Host(with network connectivity).


A simple example of SLA in practice monitoring is PING check. Ping is a network availability tool used to test the reachability of a host. When regularly sampling a host, we can save and calculate the availability of those sampled host and have a good sense of this host availability quality.


Example for ping output:


PING google.com (216.58.206.14) 56(84) bytes of data.
64 bytes from 1e120.net (216.58.206.14): icmp_seq=1 ttl=53 time=76.0 ms
64 bytes from 1e120.net (216.58.206.14): icmp_seq=2 ttl=53 time=77.2 ms
64 bytes from 1e120net (216.58.206.14): icmp_seq=3 ttl=53 time=75.0 ms
64 bytes from 1e120.net (216.58.206.14): icmp_seq=4 ttl=53 time=80.4 ms
64 bytes from 1e120.net (216.58.206.14): icmp_seq=5 ttl=53 time=97.1 ms


--- www.example.com ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4004ms
rtt min/avg/max/mdev = 75.027/81.187/97.183/8.211 ms


It's hard to imagine the world without the internet these days but not much of us really familiar with the topology of the internet, the internet is made of many networking junctions, huge data centers, millions of cables and optic fibers going on the ground and under oceans and seas, satellites and even more bizarre ways to deliver cheap or even free connectivity to far places.


When measuring the availability rate of essential and international services relying on servers, cloud servers, virtual machines and many other types of hosts with some purpose, its also very beneficial to check it from various geo locations and network junctions to be able to see the full picture and spot issues with the networking status of the monitored host. Ultimately, those checks and gathered statistics will give a broader picture of your overall service from a networking point of view.


SLA can also be implemented for other types of checks other then PING and network availiilty, those checks are often application PORTs and HTTP calls.


SLA is also ubiquitous and essential for various API health checks, so, for example, one use case can include checking the response time of some API calls such as:


https://api.yourdomain.com/webservice/clients/call1
https://api.yourdomain.com/webservice/clients/call2
https://api.yourdomain.com/webservice/clients/call3


Those URLs can be a web service, microservice, API, a website, blogs, a service, and other types of data delivery point.


Tracking the fluctuations of various simple and not invasive checks such as HTTP, PORT, PING, are the foundation and a significant matrix to consider when measuring IT services & microservices environments.


An example of SLA report taken from my Allprobe BETA account

Sample SLA report



Our October 2019 significant UI and UX update
Discovery flowchart
CMDB & IT transformation
SLA measurement high importance
CMDB made easy
The IT challenge