Microsoft Sentinel data connectors, UAE

The legacy HTTP Data Collector API stops being supported after 14 September 2026.

Microsoft has published that date and the consequence: data sources, custom integrations and connectors still using it risk ingestion interruptions afterwards. If anything you built is sending data that way, this is the most urgent Sentinel task on your list.

Microsoft Sentinel data connectors for UAE organisations
  • 14 Sep 2026Legacy HTTP Data Collector API support ends
  • 31 Mar 2027Sentinel in the Azure portal is retired
  • 4 routesFor building a custom connector
  • 3 support tiersMicrosoft, partner and community
How data gets into Sentinel

Eight things to establish about your own ingestion.

Sentinel is only as good as what reaches it, and most estates have accumulated connectors across several eras of the product. Two of those eras now have published end dates, which makes an inventory more urgent than it has been.

What is still using the legacy API

After 14 September 2026 the legacy HTTP Data Collector API is no longer supported, and Microsoft warns of potential ingestion interruptions afterwards. Anything custom built more than a couple of years ago is a candidate, and the failure mode is data quietly stopping rather than an error somebody sees.

Where that data should go instead

Microsoft recommends transitioning to the Logs Ingestion API or the Codeless Connector Framework, citing uninterrupted ingestion, improved reliability, scalability and long-term support. The choice between them usually depends on whether you are calling an API or receiving a push.

Service to service connectors for Microsoft and AWS

Out of the box service-to-service support integrates in real time. The Defender XDR connector alone brings in data from Office 365, Microsoft Entra ID, Microsoft Defender for Identity and Microsoft Defender for Cloud Apps, which is a substantial amount of coverage from one connection.

Syslog and CEF through the Azure Monitor Agent

Events stream from Linux-based Syslog-supporting devices using the Azure Monitor Agent, installed either on the device or on a dedicated Linux log forwarder. Syslog messages land in the Syslog table and CEF messages in the CommonSecurityLog table, which is worth knowing when writing queries.

Custom logs and custom tables

For some sources you collect logs as files on Windows or Linux machines, and after configuration the data appears in custom tables. That is the route for applications with no API and no Syslog output, which describes a large share of line of business software.

Who supports each connector you rely on

Connectors are Microsoft-supported, partner-supported or community-supported. Community-supported connectors have no listed support contact, and issues are raised in the Microsoft Sentinel GitHub community. Knowing which tier a critical connector sits in matters before it breaks, not after.

Connectors arrive as part of solutions

Deploying a solution from the Content Hub brings the data connector together with related workbooks, analytics rules and playbooks in the same deployment. Installing a connector alone and then wondering why there is no detection content is a common and avoidable step.

The portal move, with its own date

After 31 March 2027, Sentinel is no longer supported in the Azure portal and is available only in the Microsoft Defender portal, with customers redirected. That affects runbooks, saved links, documentation and any automation that assumes the Azure portal experience.

The near-term deadline

After 14 September 2026, anything on the legacy HTTP Data Collector API may simply stop arriving.

Microsoft frames it as potential ingestion interruptions, and the reason it deserves attention now is how it fails.

  • Ingestion stopping is silent. A connector that has worked for three years does not raise an alert when it stops sending, and a Sentinel workspace with one missing source looks identical to one with all its sources until somebody queries a table and finds nothing recent.
  • The candidates are custom work rather than product connectors: integrations built by a previous supplier, scripts pushing data from a line of business system, and anything written against the older Log Analytics ingestion pattern. Those are exactly the ones nobody has documentation for.
  • The recommended destinations are the Logs Ingestion API and the Codeless Connector Framework, and Microsoft cites uninterrupted ingestion, improved reliability, scalability and long-term support as the reasons. For most cases the Codeless Connector Framework is the lower-effort route where the source exposes an API.
  • The practical action is an inventory rather than a migration project. Establish which sources use the legacy API, which are actually still needed, and which can be retired instead of migrated. That last category is usually larger than expected.
Ask us to inventory your connectors
How we approach it

Four things that make ingestion reliable rather than assumed.

A SIEM with a missing source is worse than one you know is incomplete, because it produces confident conclusions from partial data. Almost all of this work is about making incompleteness visible.

We find the legacy API usage before the deadline finds it

After 14 September 2026 the legacy HTTP Data Collector API is unsupported and Microsoft warns of potential ingestion interruptions. The custom integrations using it are rarely documented, so this is an inventory exercise rather than a lookup, and it is worth doing now rather than in September.

We ask what actually uses each source

Sources mapped to the analytics rules and workbooks that consume them, because a source nothing queries is cost without benefit. Migrating it is wasted effort and retiring it reduces both the migration workload and the ongoing ingestion volume.

We add alerting for silence

The characteristic failure of a SIEM connector is that it stops without complaining. Alerting when a source goes quiet turns that into a detected event, and it is the single change that most improves confidence in the data underneath every detection you run.

We install solutions rather than bare connectors

Deploying a solution from the Content Hub brings the connector together with related workbooks, analytics rules and playbooks in one deployment. Connecting a source and then building detection from scratch is a common way to spend weeks recreating content that shipped with the connector.

How an engagement runs

Four phases across roughly four to six weeks.

The inventory usually finds three categories: sources to migrate, sources to retire, and sources everybody assumed were connected and are not.
  1. 01
    Week 1

    Inventory every ingestion path

    Not just the connectors page. Custom integrations, scripts, forwarders and anything a previous supplier built. Each one attributed to a source, an owner, an ingestion method and a support tier, since community-supported connectors carry a different risk profile.

    • Every ingestion path listed with method and owner
    • Legacy HTTP Data Collector API usage identified
    • Support tier recorded per connector
    • Sources believed connected but silent identified
  2. 02
    Week 2

    Decide migrate or retire, per source

    The most valuable question in the engagement. A source that nothing queries and no analytics rule uses does not need migrating, it needs turning off. That decision reduces both the migration effort and the ongoing ingestion volume.

    • Sources mapped to analytics rules and workbooks that use them
    • Unused sources identified for retirement
    • Migration target chosen per remaining source
    • Retirement decisions recorded with owners
  3. 03
    Weeks 3 to 4

    Migrate, preferring codeless where it fits

    Codeless Connector Framework where the source exposes a suitable API, Logs Ingestion API through an Azure Function or Logic App where custom logic is genuinely needed. Each migration validated by confirming data lands in the expected table before the old path is removed.

    • Connectors rebuilt on supported routes
    • Data confirmed arriving in the expected tables
    • Analytics rules validated against the new data
    • Legacy paths decommissioned only after validation
  4. 04
    Weeks 5 to 6

    Add ingestion monitoring and plan the portal move

    The step that prevents this recurring: alerting when a source stops sending, so silent ingestion failure becomes a detected event. Alongside it, planning for the Azure portal retirement on 31 March 2027, which affects runbooks and saved links.

    • Ingestion health monitoring per source
    • Alerting when a source goes quiet
    • Runbooks updated for the Defender portal
    • Content Hub solutions installed where connectors were standalone
Where this matters

Six situations where ingestion is the problem.

The symptom is usually a detection that should have fired, or an investigation that hit a gap in the data at exactly the wrong moment.

An organisation with integrations built by a previous supplier

Custom ingestion written by somebody no longer engaged, undocumented, and quite possibly using the HTTP Data Collector API that stops being supported after 14 September 2026. This is the highest priority inventory in most Sentinel estates right now.

A business that discovered a gap during an investigation

The worst way to learn a connector broke. Ingestion health monitoring with alerting on silence converts that from an incident-time discovery into a Tuesday morning ticket, which is a substantially better outcome for the same underlying failure.

A regulated firm evidencing log coverage

Auditors ask which sources feed the SIEM and how you know they are still feeding it. An inventory with owners, methods and support tiers answers the first, and ingestion health alerting answers the second in a way a screenshot cannot.

An operator with network and appliance logs

Syslog and CEF from Linux-based devices via the Azure Monitor Agent, either on the device or on a dedicated log forwarder. Knowing that Syslog lands in the Syslog table and CEF in CommonSecurityLog is also what makes the resulting queries work first time.

A company with line of business applications that only write files

No API, no Syslog, just log files on a server. Custom log collection puts that data into custom tables, which is frequently the only route for the application that matters most to the business and integrates with nothing.

An organisation planning data lake retention

The data lake tier behaves differently from the analytics tier and it is worth knowing before ingesting. Specific records cannot be purged from it, it retains data for the defined retention period even if deleted at source, and Purview setting changes do not affect it.

Three positions

How UAE organisations manage Sentinel ingestion.

The right column is more common than it should be, and its defining feature is that nobody knows what stopped working until they need the data.
Every ingestion path known
Inventoried and monitoredYes
Connected, not monitoredMostly
Accumulated over yearsNo
Legacy API usage identified
Inventoried and monitoredYes
Connected, not monitoredUnknown
Accumulated over yearsUnknown
Silent ingestion failure detected
Inventoried and monitoredAlerted
Connected, not monitoredNo
Accumulated over yearsNo
Unused sources retired
Inventoried and monitoredYes
Connected, not monitoredNo
Accumulated over yearsNo
Connectors installed with solutions
Inventoried and monitoredYes
Connected, not monitoredSometimes
Accumulated over yearsRarely
Support tier known per connector
Inventoried and monitoredYes
Connected, not monitoredNo
Accumulated over yearsNo
Sources mapped to analytics rules
Inventoried and monitoredYes
Connected, not monitoredNo
Accumulated over yearsNo
Ready for the September 2026 deadline
Inventoried and monitoredYes
Connected, not monitoredUnknown
Accumulated over yearsUnknown
Ready for the Defender portal move
Inventoried and monitoredPlanned
Connected, not monitoredNot started
Accumulated over yearsNot started
Ingestion volume understood
Inventoried and monitoredYes
Connected, not monitoredPartly
Accumulated over yearsNo
Choosing a route

How to get a source into Sentinel, and when each fits.

Most estates end up using several of these. The mistake is defaulting to whichever one the last person used rather than choosing per source.
RouteBest suited to
Solution from the Content HubAnything with an existing connector, since it brings detection content too
Service to service connectorMicrosoft services and Amazon Web Services, integrating in real time
Codeless Connector FrameworkA source that exposes an API you can poll, with no code to maintain
Logs Ingestion API in an Azure FunctionCustom logic, transformation, or authentication the framework cannot express
Logs Ingestion API in a Logic AppLow code integration where an Azure Function is more than you need
Syslog via Azure Monitor AgentNetwork and Linux devices, landing in the Syslog table
CEF via Azure Monitor AgentSecurity appliances emitting CEF, landing in CommonSecurityLog
Custom logs via the agentApplications that only write log files, landing in custom tables
Azure Monitor Agent directly or LogstashCustom connector scenarios outside the above
Legacy HTTP Data Collector APINothing, after 14 September 2026
How an engagement runs

Five steps, and the second one saves the most work.

Deciding what not to migrate is the highest leverage decision available, and it requires knowing what actually consumes each source.
  1. 1

    Inventory every path data takes into the workspace

    Product connectors, service to service integrations, agents, forwarders, scripts and anything a previous supplier built. Each attributed to a source, an owner, an ingestion method and a support tier, since community-supported connectors carry no listed support contact.

  2. 2

    Map sources to what consumes them

    Analytics rules, workbooks and hunting queries. A source nothing consumes is ingestion cost with no security benefit, and retiring it is both cheaper and faster than migrating it to a supported ingestion route.

  3. 3

    Migrate the legacy paths to supported routes

    Codeless Connector Framework where the source exposes a suitable API, or the Logs Ingestion API through an Azure Function or Logic App where custom logic is needed. Validated by confirming data lands in the expected table before the legacy path is removed.

  4. 4

    Add ingestion health monitoring

    Alerting when a source stops sending, per source rather than in aggregate. This is what turns the characteristic silent failure of SIEM ingestion into something detected, and it protects every detection built on top of that data.

  5. 5

    Plan the Defender portal transition

    After 31 March 2027 Sentinel is available only in the Microsoft Defender portal, with Azure portal users redirected. Runbooks, saved links, training material and any automation assuming the Azure portal all need updating, and doing it early avoids a rushed change.

Straight answers

What organisations ask about Sentinel data connectors.

Ingestion review

Fifteen questions about your own Sentinel data.

The first group has a deadline attached. The third is the one that determines whether you find out about a broken connector before an incident does.

Deadlines

  • Does anything use the HTTP Data Collector API?
    Support ends 14 September 2026.
  • Did a previous supplier build integrations?
    Those are the likeliest candidates.
  • Are we still using the Azure portal?
    Retired for Sentinel 31 March 2027.
  • Do our runbooks link to the Azure portal?
    They will need updating.
  • Have we planned the Defender portal move?
    There is a documented path.

Coverage

  • Which sources are connected today?
    From evidence, not memory.
  • Is the Defender XDR connector enabled?
    It brings four sources at once.
  • Are connectors installed with their solutions?
    Solutions include detection content.
  • What support tier is each connector?
    Community means no support contact.
  • Which sources have no analytics rules?
    Ingestion with no detection.

Health

  • Would we notice if a source stopped?
    Usually not.
  • Do we alert on ingestion gaps?
    The fix for the above.
  • When did each table last receive data?
    A one minute query.
  • Do CEF sources land in CommonSecurityLog?
    Syslog goes to the Syslog table.
  • Who owns connector health?
    Name them.
Next step

For each table you rely on, check when it last received data.

Tell us about source systems, required events, workspaces, network paths and data owners. We will confirm the requirements and proposed scope with you.