แก้ไข

แชร์ผ่าน

Restore logs in Azure Monitor

  • บทความ

The restore operation makes a specific time range of data in a table available in the hot cache for high-performance queries. This article describes how to restore data, query that data, and then dismiss the data when you're done.

Note

Tables with the Auxiliary table plan do not support data restore. Use a search job to retrieve data that's in long-term retention from an Auxiliary table.

Permissions

To restore data from long-term retention, you need Microsoft.OperationalInsights/workspaces/tables/write and Microsoft.OperationalInsights/workspaces/restoreLogs/write permissions to the Log Analytics workspace, for example, as provided by the Log Analytics Contributor built-in role.

When to restore logs

Use the restore operation to query data in long-term retention. You can also use the restore operation to run powerful queries within a specific time range on any Analytics table when the log queries you run on the source table can't complete within the log query timeout of 10 minutes.

Note

Restore is one method for accessing data in long-term retention. Use restore to run queries against a set of data within a particular time range. Use Search jobs to access data based on specific criteria.

What does restore do?

When you restore data, you specify the source table that contains the data you want to query and the name of the new destination table to be created.

The restore operation creates the restore table and allocates extra compute resources for querying the restored data using high-performance queries that support full KQL.

The destination table provides a view of the underlying source data, but doesn't affect it in any way. The table has no retention setting, and you must explicitly dismiss the restored data when you no longer need it.

Restore data

To restore data from a table, call the Tables - Create or Update API. The name of the destination table must end with _RST.

PUT https://management.azure.com/subscriptions/{subscriptionId}/resourcegroups/{resourceGroupName}/providers/Microsoft.OperationalInsights/workspaces/{workspaceName}/tables/{user defined name}_RST?api-version=2021-12-01-preview

Request body

The body of the request must include the following values:

Name Type Description
properties.restoredLogs.sourceTable string Table with the data to restore.
properties.restoredLogs.startRestoreTime string Start of the time range to restore.
properties.restoredLogs.endRestoreTime string End of the time range to restore.

Restore table status

The provisioningState property indicates the current state of the restore table operation. The API returns this property when you start the restore, and you can retrieve this property later using a GET operation on the table. The provisioningState property has one of the following values:

Value Description
Updating Restore operation in progress.
Succeeded Restore operation completed.
Deleting Deleting the restored table.

Sample request

This sample restores data from the month of January 2020 from the Usage table to a table called Usage_RST.

Request

PUT https://management.azure.com/subscriptions/00000000-0000-0000-0000-00000000000/resourcegroups/testRG/providers/Microsoft.OperationalInsights/workspaces/testWS/tables/Usage_RST?api-version=2021-12-01-preview

Request body:

{
    "properties":  {
    "restoredLogs":  {
                      "startRestoreTime":  "2020-01-01T00:00:00Z",
                      "endRestoreTime":  "2020-01-31T00:00:00Z",
                      "sourceTable":  "Usage"
    }
  }
}

Query restored data

Restored logs retain their original timestamps. When you run a query on restored logs, set the query time range based on when the data was originally generated.

Set the query time range by either:

  • Selecting Custom in the Time range dropdown at the top of the query editor and setting From and To values.

    or

  • Specifying the time range in the query. For example:

    let startTime =datetime(01/01/2022 8:00:00 PM);
let endTime =datetime(01/05/2022 8:00:00 PM);
TableName_RST
| where TimeGenerated between(startTime .. endTime)

Dismiss restored data

To save costs, we recommend you delete the restored table to dismiss restored data when you no longer need it.

Deleting the restored table doesn't delete the data in the source table.

Note

Restored data is available as long as the underlying source data is available. When you delete the source table from the workspace or when the source table's retention period ends, the data is dismissed from the restored table. However, the empty table will remain if you do not delete it explicitly.

Limitations

Restore is subject to the following limitations.

You can:

  • Restore data from a period of at least two days.

  • Restore up to 60 TB.

  • Run up to two restore processes in a workspace concurrently.

  • Run only one active restore on a specific table at a given time. Executing a second restore on a table that already has an active restore fails.

  • Perform up to four restores per table per week.

Pricing model

The charge for restored logs is based on the volume of data you restore, and the duration for which the restore is active. Thus, the units of price are per GB per day. Data restores are billed on each UTC-day that the restore is active.

  • Charges are subject to a minimum restored data volume of 2 TB per restore. If you restore less data, you will be charged for the 2 TB minimum each day until the restore is dismissed.

  • On the first and last days that the restore is active, you're only billed for the part of the day the restore was active.

  • The minimum charge is for a 12-hour restore duration, even if the restore is active for less than 12-hours.

  • For more information on your data restore price, see Azure Monitor pricing on the Logs tab.

Here are some examples to illustrate data restore cost calculations:

  1. If your table holds 500 GB a day and you restore 10 days data from that table, your total restore size is 5 TB. You are charged for this 5 TB of restored data each day until you dismiss the restored data. Your daily cost is 5,000 GB multiplied by your data restore price (see Azure Monitor pricing.)

  2. If instead, only 700 GB of data is restored, each day that the restore is active is billed for the 2 TB minimum restore level. Your daily cost is 2,000 GB multiplied by your data restore price.

  3. If a 5 TB data restore is only kept active for 1 hour, it is billed for 12-hour minimum. The cost for this data restore is 5,000 GB multiplied by your data restore price multiplied by 0.5 days (the 12-hour minimum).

  4. If a 700 GB data restore is only kept active for 1 hour, it is billed for 12-hour minimum. The cost for this data restore is 2,000 GB (the minimum billed restore size) multiplied by your data restore price multiplied by 0.5 days (the 12-hour minimum).

Note

There is no charge for querying restored logs since they are Analytics logs.

Next steps