> For the complete documentation index, see [llms.txt](https://docs.cow.tech/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cow.tech/using-cow/navigation/integrations/azure-table.md).

# Azure Table

Connect CoW to an **Azure Table Storage** table containing your devices — the self-serve option for technical MSPs who want to push and pull data programmatically. You run the table in your own Azure tenant and populate it however you like (script, Rewst, Power Automate, PSA export). CoW reads your devices, looks up warranties, and — optionally — writes the results back into the same rows.

## Prerequisites

Your table must contain these columns (all required — the Test step fails if any are missing):

* `ClientName`: the end client each device belongs to.
* `DeviceSerial`: the device serial number (CoW's lookup key).
* `DeviceManufacturer`: the manufacturer. If blank or `Unknown`, CoW attempts to detect it from the serial.
* `DeviceName`: a friendly device name.
* `CompanyId`: your identifier for the client.
* `DeviceId`: your identifier for the device.

{% hint style="info" %}
**Optional**: `StartDate` / `EndDate` if you already hold warranty dates (CoW reads them, and overwrites on write-back). Each row's Azure `PartitionKey`/`RowKey` are used to target write-back — keep them stable.
{% endhint %}

## Import Steps

{% stepper %}
{% step %}

### Create a SAS link

Follow Microsoft's documentation to create a [shared access signature (SAS) token](https://learn.microsoft.com/en-us/azure/ai-services/translator/document-translation/how-to-guides/create-sas-tokens?tabs=Containers). Read-only sync needs **Read** (query) permission; add **Add** and **Update** as well if you plan to enable write-back. When you copy it from Azure, paste the **SAS token string** — the part after the `?` — not the full SAS URL.
{% endstep %}

{% step %}

### Enter SAS Details

Enter the Storage Account Name, Table Name, and SAS token.
{% endstep %}

{% step %}

### Test the Connection

Click Test to confirm the token authenticates, the table is readable, and all required columns are present (your table needs at least one row for the column check).
{% endstep %}

{% step %}

### Enable the Integration

Toggle on "Enable Integration". Optionally toggle "Write Back to Table" to save warranty results (`StartDate`, `EndDate`, `Warranty`, `Status`, `WarrantyJSON`) back onto your existing rows.
{% endstep %}

{% step %}

### Submit

Click Submit to save the enablement of the integration. CoW queues the first sync immediately, then re-syncs each integration roughly once a day.
{% endstep %}
{% endstepper %}

## What COW writes back

With **Write Back to Table** on, for each device where a warranty is found CoW updates that device's **existing row** (matched on its original `PartitionKey`/`RowKey`) with `StartDate`, `EndDate` (both `YYYY-MM-DD`), `Warranty` (description), `Status`, and `WarrantyJSON` (a compact JSON of all of it plus the lookup time). Unresolved devices are left untouched and retried next sync.

## Troubleshooting

* **Test fails "missing required columns"** — add the exact column names above (case-sensitive).
* **Test fails "table is empty"** — the column check needs at least one row.
* **Test fails to authenticate** — make sure you pasted the SAS **token string** (the part after the `?`), not the full SAS URL.
* **Some devices never appear** — filtered for empty/placeholder `DeviceSerial`, blank `ClientName`, undetectable manufacturer, or your plan's device cap (upgrade to raise it).
* **Warranties aren't writing back** — confirm write-back is on, the SAS has Add/Update, and rows still have the `PartitionKey`/`RowKey` CoW first read them under.

{% hint style="warning" %}
Set the SAS token's expiry well into the future — syncs stop the moment it expires, and you'll need to generate a new one.
{% endhint %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.cow.tech/using-cow/navigation/integrations/azure-table.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.