We are designing an Azure based system that needs to be redundant and able to handle regional outage. We are using application insights for all logging. Is an Application Insights instance itself regionally distributed? or will we need to design our systems to store multiple application insights keys and switch over to backup if primary AI instance suffers a regional failure? I cannot fight any disaster recovery information on Application Insights service and best practices.
For App Insights a recommendation an App Insights per Instance of your application. Ideally an app hosted in East US will have associated App Insights in East US. Then can send to Log Analytics to query across multiple instances.
Related
I am planning to use the Azure Monitor service and specifically Workspace-based Application Insights for the client's project.
Before I purpose/implement Workspace-based Application Insights, I need to share the estimated cost with the client so I checked the Azure Pricing Calculator however in Azure Pricing Calculator for Azure Monitor service they have mentioned ONLY Classic Application Insights and I am interested in Workspace-based Application Insights so if someone has used or know about pricing then please share a link here. Please share official/authenticated information ONLY.
workspace-based application insight pricing calculator
As per this MSDoc,
Since workspace-based Application Insights resources store their data in a Log Analytics workspace, the billing for data ingestion and retention is done by the workspace where the Application Insights data is located. This enables you to leverage all options of the Log Analytics pricing model, including commitment tiers
in addition to Pay-As-You-Go.
As per Azure Monitor pricing
Application Insights charge for the data they consume.
Two different types of logs can be ingested for workspace-based Application Insights data
Analytics Logs and
Basic Logs
Application Insights data absorbed into either the workspace-based or classic versions is free to keep for 90 days without any price charges.
References taken from
Azure Monitor pricing
Application insights billing
We are designing an application which will be hosted on AKS(Azure kubernetes service). The application will consist of a set of services written in asp .net core running in docker containers. I want to monitor the services as well as the containers/nodes and have the observability across the cluster. Azure monitor for containers seems to be a good solution for monitoring containers, nodes and the cluster as a whole however I want the advanced monitoring capabilities of the application insights for the asp .net core services for example application maps, live metrics streams, transaction tracing and such features. Moreover, I don't want to have overlapping solutions. Is the Azure monitor for containers able to provide all or most of these application insights features or do I have to have both solutions in order to get proper cluster monitoring and also the advanced application monitoring?
Azure Monitor for containers provide infrastructure level monitoring and basic application logs with stdout and stderr, Kubernetes events captured out of the box.
It does not provide instrumentation for your apps or distributed tracing capabilities today, which is possible with Application Insights.
If you are looking for application map & instrumentation for events metrics and logs for your app, you can use both together and it's possible to correlate data from both and create dashboards and views.
The long term road map has Azure Monitor for containers & Application insights combined offering
In this (https://learn.microsoft.com/en-us/azure/azure-monitor/insights/container-insights-overview) guide, you can find all the features of AKS monitoring. If these covers all your use cases then you don't need to install any other tool. If not, then you cover only those features which are missing.
What would happen to Application Insight when a Azure region goes down. Will Application insight still be accessible??? as we rely on telemetry data.
There is no built-in geo-replication for Application Insights. Best practice would be for you to have an instance for each of your application deployments in every region. E.g. if your app runs in East US and in West Europe, also have AppInsights in each region and log to there respectively.
If you need to have your telemetry available in cases of an regional outage as well, I would look into Continuous Export - to a storage account in a separate region.
I'm trying to find any documentation/advisory info on whether or not to use the same App Insights instance from multiple regions. I'm assuming that if I have an API App Service in useast, it's not recommended to use an App Insights instance from West region as it would add latency.
I just got the feedback from MS application insights team, the answer is no performance issue:
Application insights sends data to their backend asynchronously - so
the actual network RT time should not matter.
Also, even though the
app insights is in West Region, the 'ingest' endpoints are globally
distributed, and telemetry is always sent to the nearest available
'ingest' point.
Details are here.
For the official document, I'm asking for it(but not sure if they would have one).
Hope it helps.
I am working on a micro-service application build in MS Azure Service Fabric. I need to gather infra level telemetry of the cluster and its nodes in real time and create Power BI queries so that i can prepare charts for the telemetry in Power BI Dashboard. I cannot use OMS due to its high cost. I was hoping to use Elastic Search but any other tools if also fine.
Thanks in advance.
Application Insights will only log information about the application. I need to capture Infrastructure related details here. Details like, downtime of a certain node on cluster, location of cluster a particular micro-service is hosted on, amount of traffic on different cluster, and so on