View source for Monitorization and observability

Continuous monitorization is a fundamental part for having a total control of everything that happens in production and helps us taking decisions for events in time, knowing the exact status of our application and infrastructure. We must have a constant visibility of what is happening and having alerts in case we have reaching a failure point or we are already on it.

This information MUST be useful, it is not enough with having a huge quantity of data if it doesn't offer qualitative information. Useful information must offer a good observability of our service with the finality of analyzing and taking decisions.

We must establish observability protocols, such checking data periodically at first time in the morning or during deployments.

We have to guarantee that the '''overall performance of the systems is stable''' or improves.

In case there is a failure in the application or infrastructure a good observability helps us noticing it.

We can't have hundreds of panels, we need to identify those really important metrics (KPIs) and observe them. For the rest we can monitor and configure thresholds with alerts.


-----

<span id="monitorization-sources"></span>
= Monitorization sources =

<span id="healthcheck"></span>
== Healthcheck ==

It allows to check at both the application and server level if service is up by checking an endpoint.

<span id="application-logs"></span>
== Application logs ==

We can visualize the reports of te application. Logs sources can be from the framework, consumers, servers, etc. For analyzing them we can use ELK or an [https://www.from-developer-to-cto.com/devops/monitorization-and-observability/monitorization-apms <u>APM</u>].

<span id="backend"></span>
== Backend ==

* use of memory
* number of processes
* requests time

<span id="most-important-pages-lighthouse"></span>
== Most important pages (Lighthouse) ==

monitor the most critical pages of the site: home, product and checkout

* weight
* requests made by the page

<span id="data"></span>
== Data ==

Database or any other data source must be monitorized for guaranteeing data consistency (such cart without customer) or alerts (no sales during last 10 minutes). Grafana or Prometheus allow analyzing various source types.

* performance while storing
* number requests
* slow queries
* most used queries

<span id="infrastructure"></span>
== Infrastructure ==

We must control everything that happens in the infrastructure for controlling problems related with server load, memory, storage, traffic, response latency, etc. Cloud server usually provides tools for it, such AWS Cloudwatch.

* performance &amp; latency
* memory
* disk


-----

<span id="monitorization-apms"></span>
= monitorization APMs =

Application Performance Monitoring tools allow analyzing our system completely, both in the application and infrastructure side. These services allow having a centralized tool for diagnosing the performance of all systems.

There are multiple SaaS that can be easily integrated into our systems with few configuration.

[https://www.datadoghq.com/ <u>Datadog</u>]

[[File:UserDocumentation_Knowledge_Central_KC_IT_Devops_definitions_monitorization_and_observabilityimage-20250421-065245.png|500px|center]]

[https://newrelic.com/ <u>New Relic</u>]

[[File:UserDocumentation_Knowledge_Central_KC_IT_Devops_definitions_monitorization_and_observabilityimage-20250421-065228.png|500px|center]]

[https://www.elastic.co/es/observability/application-performance-monitoring <u>Elastic APM</u>]

[[File:UserDocumentation_Knowledge_Central_KC_IT_Devops_definitions_monitorization_and_observabilityimage-20250421-065310.png|500px|center]]

[https://sentry.io/ <u>Sentry</u>]

[https://cloud.google.com/datastudio <u>Google Studio</u>]

[https://grafana.com/ <u>Grafana</u>]
 <!-- From bodyContent 1543250.mraw -->