wrapper for AI setup, compatible with Azure Functions
npm i appinsights-logger --save
Just set ENV variable APPINSIGHTS_INSTRUMENTATIONKEY
Set the AI_MESSAGE_NAMESPACE ENV variable to identify message source. Format is [myCompany].[AgileTeam].[Repo] (foocompany.Architecture.Appinsights)
const { trackEvent, trackException } = require('appinsights-logger')
trackEvent({ name: 'TEST', properties: { myProp: 'my value' } })
const ex = new Error('some error')
trackException({ exception: ex, properties: { myProp: 'my value' } })
import { markDependency, measureDependency } from 'appinsights-logger'
const marker = markDependency('HTTP', 'GET /products')
try {
const url = 'https://domain.com/api/products'
const locations = await requestPromise(url)
measureDependency(marker, url)
} catch(ex) {
measureDependency(marker, url, false)
// could also trackException(ex) here
}
dependency tracking is useful for measuring latency of remote calls (SQL, HTTP) and also failures
import { markDependency, measureDependency } from 'appinsights-logger'
const marker = markDependency('cosmos', 'query collection')
try {
const sql = 'select top 100 from c'
const locations = await cosmosDb.items.query(sql) // pseudo-code cosmos query
measureDependency(marker, sql)
} catch(ex) {
measureDependency(marker, sql, false)
// could also trackException(ex) here
}
customEvents are powerful because you can query the logs by the
customDimensions
object
import { trackDebugEvent } from from 'appinsights-logger'
// This 'debug' event will only log if the DEBUG_INSIGHTS="true" env var is set
trackDebugEvent({ name: 'new order', properties: { order }, measurements: { productCount: order.products.length } })
cp .env.example .env
fill in
.env
with App Insight key
npm run build
npm test