SQL Server back up to URL for Microsoft Azure Blob Storage best practices and troubleshooting
Applies to: SQL Server Azure SQL Managed Instance
This article includes best practices and troubleshooting tips for SQL Server backup and restores to Microsoft Azure Blob Storage.
For more information about using Azure Blob Storage for SQL Server backup or restore operations, see:
SQL Server Backup and Restore with Microsoft Azure Blob Storage
Tutorial: SQL Server Backup and Restore to Azure Blob Storage
The following list includes general recommendations to manage backups:
Unique file name for every backup is recommended to prevent accidentally overwriting the blobs.
When creating a container, it is recommended that you set the access level to private, so only users or accounts that can provide the required authentication information can read or write the blobs in the container.
For SQL Server databases on an instance of SQL Server running in an Azure Virtual Machine, use a storage account in the same region as the virtual machine to avoid data transfer costs between regions. Using the same region also ensures optimal performance for backup and restore operations.
Failed backup activity can result in an invalid backup file. We recommend periodic identification of failed backups and deleting the blob files. For more information, see Deleting Backup Blob Files with Active Leases
WITH COMPRESSIONoption during backup can minimize your storage costs and storage transaction costs. It can also decrease the time taken to complete the backup process.
BLOCKSIZEarguments as recommended at SQL Server Backup to URL.
SQL Server is agnostic to the type of storage redundancy used. Backup to page blobs and block blobs is supported for every storage redundancy (LRS\ZRS\GRS\RA-GRS\RA-GZRS\etc.).
Handling Large Files
The SQL Server backup operation uses multiple threads to optimize data transfer to Azure Blob Storage. However the performance depends on various factors, such as ISV bandwidth and size of the database. If you plan to back up large databases or filegroups from an on-premises SQL Server database, it is recommended that you do some throughput testing first. Azure SLA for Storage has maximum processing times for blobs that you can take into consideration.
WITH COMPRESSIONoption as recommended in the Managing Backups section, it is very important when backing up large files.
Troubleshooting backup to or restore from URL
Following are some quick ways to troubleshoot errors when backing up to or restoring from the Azure Blob Storage.
To avoid errors due to unsupported options or limitations, review the list of limitations, and support for BACKUP and RESTORE commands information in the SQL Server Backup and Restore with Microsoft Azure Blob Storage article.
Parallel backups to the same blob cause one of the backups to fail with an Initialization failed error.
In SQL Server 2016 (13.x) and later versions, block blob is preferred for Backup to URL.
If you're using page blobs with BACKUP TO URL, you can use Trace Flag 3051 to turn on logging to a specific error log with the following format in:
\<action>is one of the following:
You can also find information by reviewing the Windows Event Viewer, under Application logs with the name
The request could not be performed because of an I/O device error.
Consider COMPRESSION, MAXTRANSFERSIZE, BLOCKSIZE, and multiple URL arguments when backing up large databases. See Backing up a VLDB to Azure Blob Storage
Msg 3202, Level 16, State 1, Line 1 Write on "https://mystorage.blob.core.windows.net/mycontainer/TestDbBackupSetNumber2_0.bak" failed: 1117(The request could not be performed because of an I/O device error.) Msg 3013, Level 16, State 1, Line 1 BACKUP DATABASE is terminating abnormally.
An example resolution:
BACKUP DATABASE TestDb TO URL = 'https://mystorage.blob.core.windows.net/mycontainer/TestDbBackupSetNumber2_0.bak', URL = 'https://mystorage.blob.core.windows.net/mycontainer/TestDbBackupSetNumber2_1.bak', URL = 'https://mystorage.blob.core.windows.net/mycontainer/TestDbBackupSetNumber2_2.bak' WITH COMPRESSION, MAXTRANSFERSIZE = 4194304, BLOCKSIZE = 65536;
Message Filemark on device is not aligned.
When restoring from a compressed backup, you might see the following error:
SqlException 3284 occurred. Severity: 16 State: 5 Message Filemark on device 'https://mystorage.blob.core.windows.net/mycontainer/TestDbBackupSetNumber2_0.bak' is not aligned. Reissue the Restore statement with the same block size used to create the backupset: '65536' looks like a possible value.
To solve this error, reissue the RESTORE statement with BLOCKSIZE = 65536 specified.
Failed backup activity can result in blobs with active leases.
Error during backup due to blobs that have active lease on them:
Failed backup activity can result in blobs with active leases.
If a backup statement is reattempted, backup operation might fail with an error similar to the following:
Backup to URL received an exception from the remote endpoint. Exception Message: The remote server returned an error: (412) There is currently a lease on the blob and no lease ID was specified in the request.
If a restore statement is attempted on a backup blob file that has an active lease, the restore operation fails with an error similar to the following:
Exception Message: The remote server returned an error: (409) Conflict..
When such error occurs, the blob files need to be deleted. For more information on this scenario and how to correct this problem, see Deleting Backup Blob Files with Active Leases
OS error 50: The request is not supported
When backing up a database, you may see error
Operating system error 50(The request is not supported) for the following reasons:
- The specified storage account is not General Purpose V1/V2.
- The SAS token had a
?symbol at the beginning of the token when the credential was created. If yes, then remove it.
- The current connection is unable to connect to the storage account from the current machine using Storage Explorer or SQL Server Management Studio (SSMS).
- The policy assigned to the SAS token is expired. Create a new policy using Azure Storage Explorer and either create a new SAS token using the policy or alter the credential and try backing up again.
WITH CREDENTIAL is a new option and required to back up to or restore from the Azure Blob Storage.
Failures related to credential could be the following:
The credential specified in the **BACKUP** or **RESTORE** command does not exist.
To avoid this issue, you can include T-SQL statements to create the credential if one does not exist in the backup statement. The following is an example you can use:
IF NOT EXISTS (SELECT * FROM sys.credentials WHERE credential_identity = 'mycredential') CREATE CREDENTIAL [<credential name>] WITH IDENTITY = 'mystorageaccount' , SECRET = '<storage access key>' ;
The credential exists but the login account that is used to run the backup command does not have permissions to access the credentials. Use a login account in the db_backupoperator role with Alter any credential permissions.
Verify the storage account name and key values. The information stored in the credential must match the property values of the Azure storage account you are using in the backup and restore operations.
400 (Bad Request) errors
Using SQL Server 2012 you may encounter an error performing a backup similar to the following:
Backup to URL received an exception from the remote endpoint. Exception Message: The remote server returned an error: (400) Bad Request..
This is caused by the TLS version supported by the Azure Storage Account. Changing the supported TLS version or using the workaround listed in KB4017023.
If you are using Proxy Servers to access the internet, you may see the following issues:
Connection throttling by Proxy Servers
Proxy Servers can have settings that limit the number of connections per minute. The Backup to URL process is a multi-threaded process and hence can go over this limit. If this happens, the proxy server kills the connection. To resolve this issue, change the proxy settings so SQL Server is not using the proxy. Following are some examples of the types or error messages you may see in the error log:
Write on "https://storageaccount.blob.core.windows.net/container/BackupAzurefile.bak" failed: Backup to URL received an exception from the remote endpoint. Exception Message: Unable to read data from the transport connection: The connection was closed.
A nonrecoverable I/O error occurred on file "https://storageaccount.blob.core.windows.net/container/BackupAzurefile.bak:" Error could not be gathered from Remote Endpoint. Msg 3013, Level 16, State 1, Line 2 BACKUP DATABASE is terminating abnormally.
BackupIoRequest::ReportIoError: write failure on backup device https://storageaccount.blob.core.windows.net/container/BackupAzurefile.bak'. Operating system error Backup to URL received an exception from the remote endpoint. Exception Message: Unable to read data from the transport connection: The connection was closed.
If using page blobs, you turn on the verbose logging using the Trace Flag 3051, you may also see the following message in the logs:
HTTP status code 502, HTTP Status Message Proxy Error (The number of HTTP requests per minute exceeded the configured limit. Contact your ISA Server administrator.)
Default Proxy Settings not picked up:
Sometimes the default settings are not picked up causing proxy authentication errors such as:
A nonrecoverable I/O error occurred on file "https://storageaccount.blob.core.windows.net/container/BackupAzurefile.bak:" Backup to URL received an exception from the remote endpoint. Exception Message: The remote server returned an error: (407)* **Proxy Authentication Required.
To resolve this issue, create a configuration file that allows the Backup to URL process to use the default proxy settings using the following steps:
Create a configuration file named
BackuptoURL.exe.configwith the following xml content:
<?xml version ="1.0"?> <configuration> <system.net> <defaultProxy enabled="true" useDefaultCredentials="true"> <proxy usesystemdefault="true" /> </defaultProxy> </system.net> </configuration>
Place the configuration file in the Binn folder of the SQL Server Instance. For example, if my SQL Server is installed on the C drive of the machine, place the configuration file in
C:\Program Files\Microsoft SQL Server\MSSQL13.\<InstanceName>\MSSQL\Binn.
Submit and view feedback for