Configure the Windows Firewall to Allow SQL Server Access
Firewall systems help prevent unauthorized access to computer resources. If a firewall is turned on but not correctly configured, attempts to connect to SQL Server might be blocked.
To access an instance of the SQL Server through a firewall, you must configure the firewall on the computer that is running SQL Server. The firewall is a component of Microsoft Windows. You can also install a firewall from another company. This article discusses how to configure the Windows firewall, but the basic principles apply to other firewall programs.
This article provides an overview of firewall configuration and summarizes information of interest to a SQL Server administrator. For more information about the firewall and for authoritative firewall information, see the firewall documentation, such as Windows Firewall security deployment guide.
Users familiar with managing the Windows Firewall, and know which firewall settings they want to configure can move directly to the more advanced articles:
- Configure a Windows Firewall for Database Engine Access
- Configure the Windows Firewall to Allow Analysis Services Access
- Configure a Firewall for Report Server Access
Basic Firewall Information
Firewalls work by inspecting incoming packets, and comparing them against a set of rules. If the packet meets the standards dictated by the rules, then the firewall passes the packet to the TCP/IP protocol for additional processing. If the packet does not meet the standards specified by the rules, the firewall then discards the packet, and, if logging is enabled, creates an entry in the firewall logging file.
The list of allowed traffic is populated in one of the following ways:
Automatically: When a computer with a firewall enabled initiated communication, the firewall creates an entry in the list so that the response is allowed. This response is considered solicited traffic, and there is nothing that needs to be configured.
Manually: An administrator configures exceptions to the firewall. This allows either access to specified programs or ports on your computer. In this case, the computer accepts unsolicited incoming traffic when acting as a server, a listener, or a peer. This is the type of configuration that must be completed to connect to SQL Server.
Choosing a firewall strategy is more complex than just deciding if a given port should be open or closed. When designing a firewall strategy for your enterprise, make sure that you consider all the rules and configuration options available to you. This article does not review all the possible firewall options. We recommend that you review the following documents:
Default Firewall Settings
The first step in planning your firewall configuration is to determine the current status of the firewall for your operating system. If the operating system was upgraded from a previous version, the earlier firewall settings may have been preserved. Also, the firewall settings could have been changed by another administrator or by a Group Policy in your domain.
Turning on the firewall will affect other programs that access this computer, such as file and print sharing, and remote desktop connections. Administrators should consider all applications that are running on the computer before adjusting the firewall settings.
Programs to Configure the Firewall
Configure the Windows Firewall settings with either Microsoft Management Console or netsh.
Microsoft Management Console (MMC)
The Windows Firewall with Advanced Security MMC snap-in lets you configure more advanced firewall settings. This snap-in presents most of the firewall options in an easy-to-use manner, and presents all firewall profiles. For more information, see Using the Windows Firewall with Advanced Security Snap-in later in this article.
The netsh.exe tool can be used by an administrator to configure and monitor Windows-based computers at a command prompt or using a batch file**.** By using the netsh tool, you can direct the context commands you enter to the appropriate helper, and the helper then performs the command. A helper is a Dynamic Link Library (.dll) file that extends the functionality of the netsh tool by providing configuration, monitoring, and support for one or more services, utilities, or protocols. All operating systems that support SQL Server have a firewall helper. Windows Server 2008 also has an advanced firewall helper called advfirewall. The details of using netsh are not discussed in this article. However, many of the configuration options described can be configured by using netsh. For example, run the following script at a command prompt to open TCP port 1433:
netsh firewall set portopening protocol = TCP port = 1433 name = SQLPort mode = ENABLE scope = SUBNET profile = CURRENT
A similar example using the Windows Firewall for Advanced Security helper:
netsh advfirewall firewall add rule name = SQLPort dir = in protocol = tcp action = allow localport = 1433 remoteip = localsubnet profile = DOMAIN
For more information about netsh, see the following links:
For Linux: On Linux, you also need to open the ports associated with the services you need access to. Different distributions of Linux and different firewalls have their own procedures. For two examples, see SQL Server on Red Hat, and SQL Server on SUSE.
Ports Used By SQL Server
The following tables can help you identify the ports being used by SQL Server.
Ports Used By the Database Engine
By default, the typical ports used by SQL Server and associated database engine services are: TCP 1433, 4022, 135, 1434, UDP 1434. The table below explains these ports in greater detail. A named instance uses dynamic ports.
The following table lists the ports that are frequently used by the Database Engine.
|Default instance running over TCP||TCP port 1433||This is the most common port allowed through the firewall. It applies to routine connections to the default installation of the Database Engine, or a named instance that is the only instance running on the computer. (Named instances have special considerations. See Dynamic Ports later in this article.)|
|Named instances with default port||The TCP port is a dynamic port determined at the time the Database Engine starts.||See the discussion below in the section Dynamic Ports. UDP port 1434 might be required for the SQL Server Browser Service when you are using named instances.|
|Named instances with fixed port||The port number configured by the administrator.||See the discussion below in the section Dynamic Ports.|
|Dedicated Admin Connection||TCP port 1434 for the default instance. Other ports are used for named instances. Check the error log for the port number.||By default, remote connections to the Dedicated Administrator Connection (DAC) are not enabled. To enable remote DAC, use the Surface Area Configuration facet. For more information, see Surface Area Configuration.|
|SQL Server Browser service||UDP port 1434||The SQL Server Browser service listens for incoming connections to a named instance and provides the client the TCP port number that corresponds to that named instance. Normally the SQL Server Browser service is started whenever named instances of the Database Engine are used. The SQL Server Browser service does not have to be started if the client is configured to connect to the specific port of the named instance.|
|Instance with HTTP endpoint.||Can be specified when an HTTP endpoint is created. The default is TCP port 80 for CLEAR_PORT traffic and 443 for SSL_PORT traffic.||Used for an HTTP connection through a URL.|
|Default instance with HTTPS endpoint||TCP port 443||Used for an HTTPS connection through a URL. HTTPS is an HTTP connection that uses Transport Layer Security (TLS), previously known as Secure Sockets Layer (SSL).|
|Service Broker||TCP port 4022. To verify the port used, execute the following query:
||There is no default port for SQL ServerService Broker, but this is the conventional configuration used in Books Online examples.|
|Database Mirroring||Administrator chosen port. To determine the port, execute the following query:
||There is no default port for database mirroring however Books Online examples use TCP port 5022 or 7022. It is important to avoid interrupting an in-use mirroring endpoint, especially in high-safety mode with automatic failover. Your firewall configuration must avoid breaking quorum. For more information, see Specify a Server Network Address (Database Mirroring).|
|Replication||Replication connections to SQL Server use the typical regular Database Engine ports (TCP port 1433 for the default instance, etc.)
Web synchronization and FTP/UNC access for replication snapshot require additional ports to be opened on the firewall. To transfer initial data and schema from one location to another, replication can use FTP (TCP port 21), or sync over HTTP (TCP port 80) or File Sharing. File sharing uses UDP port 137 and 138, and TCP port 139 if it using NetBIOS. File Sharing uses TCP port 445.
|For sync over HTTP, replication uses the IIS endpoint (ports for which are configurable but is port 80 by default), but the IIS process connects to the backend SQL Server through the standard ports (1433 for the default instance.
During Web synchronization using FTP, the FTP transfer is between IIS and the SQL Server publisher, not between subscriber and IIS.
|Transact-SQL debugger||TCP port 135
See Special Considerations for Port 135
The IPsec exception might also be required.
|If using Visual Studio, on the Visual Studio host computer, you must also add Devenv.exe to the Exceptions list and open TCP port 135.
If using Management Studio, on the Management Studio host computer, you must also add ssms.exe to the Exceptions list and open TCP port 135. For more information, see Configure firewall rules before running the TSQL Debugger.
For step by step instructions to configure the Windows Firewall for the Database Engine, see Configure a Windows Firewall for Database Engine Access.
By default, named instances (including SQL Server Express) use dynamic ports. That means that every time that the Database Engine starts, it identifies an available port and uses that port number. If the named instance is the only instance of the Database Engine installed, it will probably use TCP port 1433. If other instances of the Database Engine are installed, it will probably use a different TCP port. Because the port selected might change every time that the Database Engine is started, it is difficult to configure the firewall to enable access to the correct port number. Therefore, if a firewall is used, we recommend reconfiguring the Database Engine to use the same port number every time. This is called a fixed port or a static port. For more information, see Configure a Server to Listen on a Specific TCP Port (SQL Server Configuration Manager).
An alternative to configuring a named instance to listen on a fixed port is to create an exception in the firewall for a SQL Server program such as sqlservr.exe (for the Database Engine). This can be convenient, but the port number will not appear in the Local Port column of the Inbound Rules page when you are using the Windows Firewall with Advanced Security MMC snap-in. This can make it more difficult to audit which ports are open. Another consideration is that a service pack or cumulative update can change the path to the SQL Server executable which will invalidate the firewall rule.
To add a program exception to the firewall using Windows Firewall with Advanced Security
From the start menu, type wf.msc. Select Windows Firewall with Advanced Security.
In the left pane, select Inbound rules.
In the right pane, under Actions select New rule.... New Inbound Rule Wizard opens.
On Rule type, select Program. Select Next.
On Program, select This program path. Select Browse to locate your instance of SQL Server. The program is called sqlservr.exe. It is normally located at:
C:\Program Files\Microsoft SQL Server\MSSQL13.<InstanceName>\MSSQL\Binn\sqlservr.exe
On Action, click Allow the connection.
Profile, include all three profiles. Select Next.
On Name, type a name for the rule. Select Finish.
For more information about endpoints, see Configure the Database Engine to Listen on Multiple TCP Ports and Endpoints Catalog Views (Transact-SQL).
Ports Used By Analysis Services
By default, the typical ports used by SQL Server Analysis Services and associated services are: TCP 2382, 2383, 80, 443. The table below explains these ports in greater detail.
The following table lists the ports that are frequently used by Analysis Services.
|Analysis Services||TCP port 2383 for the default instance||The standard port for the default instance of Analysis Services.|
|SQL Server Browser service||TCP port 2382 only needed for an Analysis Services named instance||Client connection requests for a named instance of Analysis Services that do not specify a port number are directed to port 2382, the port on which SQL Server Browser listens. SQL Server Browser then redirects the request to the port that the named instance uses.|
|Analysis Services configured for use through IIS/HTTP
(The PivotTable® Service uses HTTP or HTTPS)
|TCP port 80||Used for an HTTP connection through a URL.|
|Analysis Services configured for use through IIS/HTTPS
(The PivotTable® Service uses HTTP or HTTPS)
|TCP port 443||Used for an HTTPS connection through a URL. HTTPS is an HTTP connection that uses TLS.|
If users access Analysis Services through IIS and the Internet, you must open the port on which IIS is listening and specify that port in the client connection string. In this case, no ports have to be open for direct access to Analysis Services. The default port 2389, and port 2382, should be restricted together with all other ports that are not required.
For step by step instructions to configure the Windows Firewall for Analysis Services, see Configure the Windows Firewall to Allow Analysis Services Access.
Ports Used By Reporting Services
By default, the typical ports used by SQL Server Reporting SErvices and associated services are: TCP 80, 443. The table below explains these ports in greater detail.
The following table lists the ports that are frequently used by Reporting Services.
|Reporting Services Web Services||TCP port 80||Used for an HTTP connection to Reporting Services through a URL. We recommend that you do not use the preconfigured rule World Wide Web Services (HTTP). For more information, see the Interaction with Other Firewall Rules section below.|
|Reporting Services configured for use through HTTPS||TCP port 443||Used for an HTTPS connection through a URL. HTTPS is an HTTP connection that uses TLS. We recommend that you do not use the preconfigured rule Secure World Wide Web Services (HTTPS). For more information, see the Interaction with Other Firewall Rules section below.|
When Reporting Services connects to an instance of the Database Engine or Analysis Services, you must also open the appropriate ports for those services. For step-by-step instructions to configure the Windows Firewall for Reporting Services, Configure a Firewall for Report Server Access.
Ports Used By Integration Services
The following table lists the ports that are used by the Integration Services service.
|Microsoft remote procedure calls (MS RPC)
Used by the Integration Services runtime.
|TCP port 135
See Special Considerations for Port 135
|The Integration Services service uses DCOM on port 135. The Service Control Manager uses port 135 to perform tasks such as starting and stopping the Integration Services service and transmitting control requests to the running service. The port number cannot be changed.
This port is only required to be open if you are connecting to a remote instance of the Integration Services service from Management Studio or a custom application.
For step-by-step instructions to configure the Windows Firewall for Integration Services, see Integration Services Service (SSIS Service).
Additional Ports and Services
The following table lists ports and services that SQL Server might depend on.
|Windows Management Instrumentation
For more information about WMI, see WMI Provider for Configuration Management Concepts
|WMI runs as part of a shared service host with ports assigned through DCOM. WMI might be using TCP port 135.
See Special Considerations for Port 135
|SQL Server Configuration Manager uses WMI to list and manage services. We recommend that you use the preconfigured rule group Windows Management Instrumentation (WMI). For more information, see the Interaction with Other Firewall Rules section below.|
|Microsoft Distributed Transaction Coordinator (MS DTC)||TCP port 135
See Special Considerations for Port 135
|If your application uses distributed transactions, you might have to configure the firewall to allow Microsoft Distributed Transaction Coordinator (MS DTC) traffic to flow between separate MS DTC instances, and between the MS DTC and resource managers such as SQL Server. We recommend that you use the preconfigured Distributed Transaction Coordinator rule group.
When a single shared MS DTC is configured for the entire cluster in a separate resource group, you should add sqlservr.exe as an exception to the firewall.
|The browse button in Management Studio uses UDP to connect to the SQL Server Browser Service. For more information, see SQL Server Browser Service (Database Engine and SSAS).||UDP port 1434||UDP is a connectionless protocol.
The firewall has a setting (UnicastResponsesToMulticastBroadcastDisabled Property of the INetFwProfile Interface) which controls the behavior of the firewall with respect to unicast responses to a broadcast (or multicast) UDP request. It has two behaviors:
If the setting is TRUE, no unicast responses to a broadcast are permitted at all. Enumerating services will fail.
If the setting is FALSE (default), unicast responses are permitted for 3 seconds. The length of time is not configurable. In a congested or high-latency network, or for heavily loaded servers, tries to enumerate instances of SQL Server might return a partial list, which might mislead users.
|IPsec traffic||UDP port 500 and UDP port 4500||If the domain policy requires network communications to be done through IPsec, you must also add UDP port 4500 and UDP port 500 to the exception list. IPsec is an option using the New Inbound Rule Wizard in the Windows Firewall snap-in. For more information, see Using the Windows Firewall with Advanced Security Snap-in below.|
|Using Windows Authentication with Trusted Domains||Firewalls must be configured to allow authentication requests.||For more information, see How to configure a firewall for domains and trusts.|
|SQL Server and Windows Clustering||Clustering requires additional ports that are not directly related to SQL Server.||For more information, see Enable a network for cluster use.|
|URL namespaces reserved in the HTTP Server API (HTTP.SYS)||Probably TCP port 80, but can be configured to other ports. For general information, see Configuring HTTP and HTTPS.||For SQL Server specific information about reserving an HTTP.SYS endpoint using HttpCfg.exe, see About URL Reservations and Registration (SSRS Configuration Manager).|
Special Considerations for Port 135
When you use RPC with TCP/IP or with UDP/IP as the transport, inbound ports are frequently dynamically assigned to system services as required; TCP/IP and UDP/IP ports that are larger than port 1024 are used. These are frequently informally referred to as "random RPC ports." In these cases, RPC clients rely on the RPC endpoint mapper to tell them which dynamic ports were assigned to the server. For some RPC-based services, you can configure a specific port instead of letting RPC assign one dynamically. You can also restrict the range of ports that RPC dynamically assigns to a small range, regardless of the service. Because port 135 is used for many services, it is frequently attacked by malicious users. When opening port 135, consider restricting the scope of the firewall rule.
For more information about port 135, see the following references:
- Service overview and network port requirements for the Windows Server system
- Troubleshooting RPC Endpoint Mapper errors using the Windows Server 2003 Support Tools from the product CD
- Remote procedure call (RPC)
- How to configure RPC dynamic port allocation to work with firewalls
Interaction with Other Firewall Rules
The Windows Firewall uses rules and rule groups to establish its configuration. Each rule or rule group is generally associated with a particular program or service, and that program or service might modify or delete that rule without your knowledge. For example, the rule groups World Wide Web Services (HTTP) and World Wide Web Services (HTTPS) are associated with IIS. Enabling those rules will open ports 80 and 443, and SQL Server features that depend on ports 80 and 443 will function if those rules are enabled. However, administrators configuring IIS might modify or disable those rules. Therefore, if you are using port 80 or port 443 for SQL Server, you should create your own rule or rule group that maintains your desired port configuration independently of the other IIS rules.
The Windows Firewall with Advanced Security MMC snap-in allows any traffic that matches any applicable allow rule. So if there are two rules that both apply to port 80 (with different parameters), traffic that matches either rule will be permitted. So if one rule allows traffic over port 80 from local subnet and one rule allows traffic from any address, the net effect is that all traffic to port 80 is permitted regardless of the source. To effectively manage access to SQL Server, administrators should periodically review all firewall rules enabled on the server.
Overview of Firewall Profiles
Firewall profiles are used by the operating systems to identify and remember each of the networks to which they connect with regard to connectivity, connections, and category.
There are three network location types in Windows Firewall with Advanced Security:
- Domain: Windows can authenticate access to the domain controller for the domain to which the computer is joined.
- Public: Other than domain networks, all networks are initially categorized as public. Networks that represent direct connections to the Internet or are in public locations, such as airports and coffee shops should be left public.
- Private: A network identified by a user or application as private. Only trusted networks should be identified as private networks. Users will likely want to identify home or small business networks as private.
The administrator can create a profile for each network location type, with each profile containing different firewall policies. Only one profile is applied at any time. Profile order is applied as follows:
- If all interfaces are authenticated to the domain controller for the domain of which the computer is a member, the domain profile is applied.
- If all interfaces are either authenticated to the domain controller or are connected to networks that are classified as private network locations, the private profile is applied.
- Otherwise, the public profile is applied.
Use the Windows Firewall with Advanced Security MMC snap-in to view and configure all firewall profiles. The Windows Firewall item in Control Panel only configures the current profile.
Additional Firewall Settings Using the Windows Firewall Item in Control Panel
Exceptions that you add to the firewall can restrict the opening of the port to incoming connections from specific computers or the local subnet. This restriction of the scope of the port opening can reduce how much your computer is exposed to malicious users, and is recommended.
Using the Windows Firewall item in Control Panel only configures the current firewall profile.
To change the scope of a firewall exception using the Windows Firewall item in Control Panel
In the Windows Firewall item in Control Panel, select a program or port on the Exceptions tab, and then click Properties or Edit.
In the Edit a Program or Edit a Port dialog box, click Change Scope.
Choose one of the following options:
Any computer (including those on the Internet): Not recommended. This will allow any computer that can address your computer to connect to the specified program or port. This setting might be necessary to allow information to be presented to anonymous users on the internet, but increases your exposure to malicious users. Your exposure can be further increased if you enable this setting and also allow Network Address Translation (NAT) traversal, such as the Allow edge traversal option.
My network (subnet) only: This is a more secure setting than Any computer. Only computers on the local subnet of your network can connect to the program or port.
Custom list: Only computers that have the IP addresses listed can connect. This can be a more secure setting than My network (subnet) only, however, client computers using DHCP can occasionally change their IP address. Then the intended computer will not be able to connect. Another computer, which you had not intended to authorize, might accept the listed IP address and then be able to connect. The Custom list option might be appropriate for listing other servers that are configured to use a fixed IP address; however, IP addresses might be spoofed by an intruder. Restricting firewall rules are only as strong as your network infrastructure.
Using the Windows Firewall with Advanced Security Snap-in
Additional advanced firewall settings can be configured by using the Windows Firewall with Advanced Security MMC snap-in. The snap-in includes a rule wizard and exposes additional settings that are not available in the Windows Firewall item in Control Panel. These settings include the following:
- Encryption settings
- Services restrictions
- Restricting connections for computers by name
- Restricting connections to specific users or profiles
- Edge traversal allowing traffic to bypass Network Address Translation (NAT) routers
- Configuring outbound rules
- Configuring security rules
- Requiring IPsec for incoming connections
To create a new firewall rule using the New Rule wizard
- On the Start menu, select Run, type WF.msc, and then select OK.
- In the Windows Firewall with Advanced Security, in the left pane, right-click Inbound Rules, and then select New Rule.
- Complete the New Inbound Rule Wizard using the settings that you want.
Troubleshooting Firewall Settings
The following tools and techniques can be useful in troubleshooting firewall issues:
The effective port status is the union of all rules related to the port. When trying to block access through a port, it can be helpful to review all the rules that cite the port number. To do this, use the Windows Firewall with Advanced Security MMC snap-in and sort the inbound and outbound rules by port number.
Review the ports that are active on the computer on which SQL Server is running. This review process includes verifying which TCP/IP ports are listening and also verifying the status of the ports.
To verify which ports are listening, use the netstat command-line utility. In addition to displaying active TCP connections, the netstat utility also displays a variety of IP statistics and information.
To list which TCP/IP ports are listening
Open the Command Prompt window.
At the command prompt, type netstat -n -a.
The -n switch instructs netstat to numerically display the address and port number of active TCP connections. The -a switch instructs netstat to display the TCP and UDP ports on which the computer is listening.
The PortQry utility can be used to report the status of TCP/IP ports as listening, not listening, or filtered. (With a filtered status, the port might or might not be listening; this status indicates that the utility did not receive a response from the port.) The PortQry utility is available for download from the Microsoft Download Center.