Internet Server Unavailable Because of Malicious SYN Attacks (142641)


The information in this article applies to:

  • Microsoft Windows 2000 Server
  • Microsoft Windows 2000 Advanced Server
  • Microsoft Windows 2000 Professional
  • Microsoft Windows NT Server 3.51
  • Microsoft Windows NT Server 4.0
This article was previously published under Q142641

SYMPTOMS

On your computer running the TCP/IP protocol and connected to the Internet, some or all network services are rendered unavailable and error messages such as the following appear on the network client screen:
The connection has been reset by the remote host.

This symptom of all network services being rendered unavailable may also occur on a computer running an operating system other than Windows NT, for example, UNIX.

CAUSE

Your computer has become the target of a malicious attack known as TCP/IP "SYN Flooding" or "SYN Attacks."

"Computer hackers" can target an entire machine, or a specific TCP service such as web services. The attack is focused on the TCP protocol used by all computers on the Internet, and is not specific to the Windows NT operating system.

How SYN Flooding Works

SYN Flooding works as follows: (see also CERT(sm) Advisory CA-96.21 at ftp://info.cert.org/pub/cert_advisories
  • A TCP connection request (SYN) is sent to the target computer. The source IP address in the packet is "spoofed," or replaced with an address that is not in use on the Internet, or that belongs to another computer. An attacker will send many of these TCP SYNs to tie up as many resources as possible on the target computer.
  • Upon receiving the connection request, the target computer allocates resources to handle and track the new connection, then responds with a "SYN-ACK". In this case, the response is sent to the "spoofed" non- existent IP address.
  • No response is received to the SYN-ACK. A default-configured Windows NT 3.5x or 4.0 computer will retransmit the SYN-ACK 5 times, doubling the time-out value after each retransmission. The initial time-out value is three seconds, so retries are attempted at 3, 6, 12, 24, and 48 seconds. After the last retransmission, 96 seconds are allowed to pass before the computer gives up on receiving a response, and deallocates the resources that were set aside earlier for the connection. The total elapsed time that resources are in use is 189 seconds.

How to Verify Your Computer is Under a SYN Attack

If you suspect that your computer is the target of a SYN attack, you can type the following command at a command prompt to view connections in the "SYN_RECEIVED" state:

netstat -n -p tcp

This command may cause the following text to appear on your screen:

Active Connections 

      Proto  Local Address         Foreign Address       State
      TCP    127.0.0.1:1030        127.0.0.1:1032        ESTABLISHED
      TCP    127.0.0.1:1032        127.0.0.1:1030        ESTABLISHED
      TCP    10.57.8.190:21        10.57.14.154:1256     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1257     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1258     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1259     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1260     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1261     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1262     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1263     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1264     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1265     SYN_RECEIVED
      TCP    10.57.8.190:21        10.57.14.154:1266     SYN_RECEIVED
      TCP    10.57.8.190:4801      10.57.14.221:139      TIME_WAIT


If a large number of connections are in the SYN_RECEIVED state, it is possible that the system is under attack. A network analyzer can be used to track the problem down further, and it may be necessary to contact your Internet Service Provider for assistance in attempting to trace the source.

The effect of tying up connection resources varies, depending upon the TCP/IP stack and applications listening on the TCP port. For most stacks, there is a limit on the number of connections that can be in the half-open (SYN_RECEIVED) state. Once the limit is reached for a given TCP port, the target computer responds with a reset to all further connection requests until resources are freed.

RESOLUTION

Obtain the following update for Windows NT 3.51 or the latest Service Pack for Windows NT 4.0

Microsoft is committed to providing the best possible protection against these attacks from within the Windows NT operating system and has made a number of changes to its TCP/IP components available to customers in response to this threat.

This global key has been designed to protect against attacks and sets the other keys mentioned later in the article to known effective values. This key avoids having the administrator guess about which values would offer the most protection. It is strongly recommended that the following global keys be used:

SynAttackProtect
Key: Tcpip\Parameters
Value Type: REG_DWORD
Valid Range: 0, 1, 2
0 (no synattack protection)
1 (reduced retransmission retries and delayed RCE (route cache entry) creation if the TcpMaxHalfOpen and TcpMaxHalfOpenRetried settings are satisfied.)
2 (in addition to 1 a delayed indication to Winsock is made.)

NOTE: When the system finds itself under attack the following options on any socket can no longer be enabled : sScalable windows (RFC 1323) and per adapter configured TCP parameters (Initial RTT, window size). This is because when protection is functioning the route cache entry is not queried before the SYN-ACK is sent and the Winsock options are not available at this stage of the connection.

Default: 0 (False)
Recommendation: 2
Description: Synattack protection involves reducing the amount of retransmissions for the SYN-ACKS, which will reduce the time for which resources have to remain allocated. The allocation of route cache entry resources is delayed until a connection is made. If synattackprotect = 2, then the connect on indication to AFD is delayed until the three-way handshake is completed. Also note that the actions taken by the protection mechanism only occur if TcpMaxHalfOpen and TcpMaxHalfOpenRetried settings are exceeded.

NOTE: The following keys should only be altered if the above global key has proven to be ineffective or specific resource limits are being hit.

The changes are listed here: 

*******************************************************************
*  1. Tcpip.sys times out half-open connections faster            *
*******************************************************************


A new version of Tcpip.sys has been produced that allows control of the number of times a response to a TCP connection request (SYN-ACK) will be retransmitted. Control is handled through a new registry parameter: 

  HKEY_LOCAL_MACHINE
  \SYSTEM
   \CurrentControlSet
    \Services
     \Tcpip
      \Parameters
       \TcpMaxConnectResponseRetransmissions
           Value Type: REG_DWORD
           Valid Range: 0-0xFFFFFFFF
           Default: 2


The default value for this parameter is now 3. The following table shows Windows NT 4.0 TCP/IP behavior for various values of this parameter: 

Value  Retransmission Times  Elapsed Time  Comments
3      3, 6, and 12 seconds  45 seconds    Cleanup 24 secs after last retx
2      3, and 6 seconds      21 seconds    Cleanup 12 secs after last retx
1      3 seconds             9  seconds    Cleanup 6  secs after last retx


This parameter changes the default time that it takes to clean up a half- open TCP connection from 189 seconds to 45 seconds, and provides more granular control to the administrator. A site that is under heavy attack might set the value as low as "1". A value of "0" is also valid; however if this parameter is set to 0, SYN-ACKs will not be retransmitted at all, and will time out in 3 seconds. With the value this low, legitimate connection attempts from distant clients may fail. 

*******************************************************************
*  2. NetBT has a Higher, Configurable Backlog                    *
*******************************************************************


NetBT (NetBIOS over TCP/IP) uses TCP port 139 and is used by Microsoft Network Services such as file and print sharing. Version 3.51 and 4.0 NetBT has a "backlog" of connection blocks available that is two plus an incremental number depending on the NetBT clients (such as the redirector, server, and any NetBIOS applications running). On a typical server, this number will be 7-11. A new version of NetBT has been produced that automatically allocates more connection blocks as needed, in a configurable manner.

On a connection event, it now checks to see if the number of free blocks is below 2, and if so, adds an "increment" number of blocks, where "increment" is configurable in the registry as shown here: 

  HKEY_LOCAL_MACHINE
  \SYSTEM
   \CurrentControlSet
    \Services
     \NetBt
      \Parameters
       \BacklogIncrement
           Value Type: REG_DWORD
           Valid Range: 1-0x14  (1-20 decimal)
           Default: 3


Each connection block consumes 78 bytes of memory. The total number of connection blocks that can be allocated by NetBT is also registry configurable: 

  HKEY_LOCAL_MACHINE
  \SYSTEM
   \CurrentControlSet
    \Services
     \NetBt
      \Parameters
       \MaxConnBackLog
           Value Type: REG_DWORD
           Valid Range: 1-0x9c40 (1-40,000 decimal)
           Default: 1000


MaxConnBackLog defaults to 1000, but can be set as high as 40,000. Connection blocks are "scavenged," or recycled, when the SYN-ACK retransmission timer expires and TCP fails the connection attempt. 

*******************************************************************
*  3. Afd.sys has been modified to withstand large numbers of     *
*     "half-open" connections efficiently                         *
*******************************************************************


Windows Sockets applications such as ftp servers and web servers have their connection attempts handled by Afd.sys. Afd.sys has been modified to support large numbers of connections in the "half-open" state without denying access to legitimate clients. This is accomplished by allowing the administrator to configure a dynamic backlog.

The new version of Afd.sys supports four new registry parameters that can be used to control the dynamic backlog behavior.

EnableDynamicBacklog is a global switch to enable or disable dynamic backlog. It defaults to 0 (off), and this setting provides no change from the existing versions. Setting it to 1 enables the new dynamic backlog feature. 

  HKEY_LOCAL_MACHINE
  \SYSTEM
   \CurrentControlSet
    \Services
     \AFD
      \Parameters
       \EnableDynamicBacklog
           Value Type: REG_DWORD
           Valid Range: 0,1
           Default: 0
           Suggested value for a system under heavy attack: 1


MinimumDynamicBacklog controls the minimum number of free connections allowed on a listening endpoint. If the number of free connections drops below this value, then a thread is queued to create additional free connections. This value should not be made too large, as the dynamic backlog code engages whenever the number of free connections falls below this value. Too large a value may lead to a performance reduction. 

  HKEY_LOCAL_MACHINE
  \SYSTEM
   \CurrentControlSet
    \Services
     \AFD
      \Parameters
       \MinimumDynamicBacklog
           Value Type: REG_DWORD
           Valid Range: 0-0xFFFFFFFF
           Default: 0
           Suggested value for a system under heavy attack: 20


MaximumDynamicBacklog controls the maximum number of "quasi-free" connections allowed on a listening endpoint. "Quasi-free" connections include the number of free connections plus those connections in a half- connected (SYN_RECEIVED) state. No attempt is made to create additional free connections if doing so would exceed this value. 

  HKEY_LOCAL_MACHINE
  \SYSTEM
   \CurrentControlSet
    \Services
     \AFD
      \Parameters
       \MaximumDynamicBacklog
           Value Type: REG_DWORD
           Valid Range: 0-0xFFFFFFFF
           Default: 0
           Suggested value for a system under heavy attack: Memory
           dependent. This value should not exceed 5000 for each 32 MB of
           RAM installed in the server, in order to prevent exhaustion of
           non-paged pool when under attack.


DynamicBacklogGrowthDelta controls the number of free connections to create when additional connections are necessary. Be careful with this value, as a large value could lead to explosive free connection allocations. 

  HKEY_LOCAL_MACHINE
  \SYSTEM
   \CurrentControlSet
    \Services
     \AFD
      \Parameters
       \DynamicBacklogGrowthDelta
           Value Type: REG_DWORD
           Valid Range: 0-0xFFFFFFFF
           Default: 0
           Suggested value for a system under heavy attack: 10 (0xa)



MaximumDynamicBacklog,



To take advantage of the changes to Afd.sys, Windows Sockets applications must specifically request a backlog greater than the value configured for MinimumDynamicBacklog when they issue their listen() call. Microsoft applications such as Internet Information Server (which has a default backlog of 25) are configurable. Application-specific details are available from the Microsoft Knowledge Base at:

http://support.microsoft.com



The modified drivers for Windows NT and instructions for installing them are available from Microsoft support channels or directly from the following Internet locations:

For Windows NT 3.51:

ftp://ftp.microsoft.com/bussys/winnt /winnt-public/fixes/usa/nt351/hotfixes-postsp5/syn-attack

STATUS

Windows NT 4.0

This problem has been corrected in the latest Microsoft Windows NT U.S. Service Pack for Windows NT 4.0. For information on obtaining the Service Pack, query on the following word in the Microsoft Knowledge Base (without the spaces):

S E R V P A C K


Windows NT 3.51

Microsoft has confirmed this problem could result in some degree of security vulnerability in Windows NT version 3.51. A fully supported fix is now available, but it has not been fully regression tested and should only be applied to systems determined to be at risk of attack. Please evaluate your system's physical accessibility, network and Internet connectivity, and other factors to determine the degree of risk to your system. If your system is sufficiently at risk, Microsoft recommends you apply this fix. Otherwise, wait for the next Windows NT service pack, which will contain this fix. Please contact Microsoft Technical Support for more information.



MORE INFORMATION

Adding these registry changes may have an adverse affect on a Microsoft Exchange Cluster. Microsoft Exchange Cluster (Exres.dll) frequently initiates connections to the SMTP, IMAP, POP3 and HTTP ports to test availability. The tests are similar to a telnet session to port 25, 143, 110, or 80. If the tests are successful, the cluster knows that the services are available to users and marks them as "Alive". If the tests are not successful, the Cluster Administrator marks the resource as offline in Cluster Administrator, and logs an event in the Application log. The event is: Event Type: Error
Event Source: MSExchangeCluster
Event Category: Services
Event ID: 2074
Date: date
Time: time
User: N/A
Computer: computer name
Description: SMTP Virtual Server Instance - (125-VS2-NAME): Cluster Service failed the isalive checking for the resource.
Modification Type:MajorLast Reviewed:5/7/2003
Keywords:kbnetwork KB142641
©2002 Microsoft Corporation. All rights reserved.