Sasser (computer worm)
View on WikipediaThis article includes a list of general references, but it lacks sufficient corresponding inline citations. (June 2025) |
| Sasser | |
|---|---|
| Malware details | |
| Technical name |
|
| Type | Worm |
| Author | Sven Jaschan |
| Technical details | |
| Platforms | Windows 2000, Windows XP |
Sasser is a computer worm that affects computers running vulnerable versions of the Windows XP and Windows 2000 operating systems. Sasser spreads by exploiting the system through a vulnerable port and can spread without user intervention. It is stopped by a properly configured firewall or by downloading system updates from Windows Update. The specific hole Sasser exploits was documented and patched by Microsoft prior to the release of the worm.
The most characteristic experience of the worm is the shutdown timer that appears due to the worm crashing LSASS. Sasser impacted various organizations including Agence France-Presse (AFP) having all its satellite communications blocked for hours and the U.S. flight company Delta Air Lines having to cancel several trans-Atlantic flights.
History
[edit]The Sasser computer worm was created on April 29, 2004.[1] The LSASS vulnerability was patched by Microsoft in the April 2004 installment of its monthly security packages,[2] prior to the release of the worm.
Behavior
[edit]The specific hole Sasser exploits is documented by Microsoft in its MS04-011 bulletin (CVE-2003-0533).[3] Sasser spreads by exploiting the system through a vulnerable port. Thus, it is particularly virulent in that it can spread without user intervention, but it is also easily stopped by a properly configured firewall or by downloading system updates from Windows Update.
The worm was named Sasser because it spreads by exploiting a buffer overflow in the component known as LSASS (Local Security Authority Subsystem Service) on the affected operating systems (vulnerable versions of the Microsoft operating systems Windows XP and Windows 2000). This buffer overflow gives a long string to an undocumented API in Microsoft Active Directory-related functions, which both allows for arbitrary code execution and often crashes LSASS.exe.[4]
Once on a machine, the worm scans different ranges of IP addresses and connects to victims' computers primarily through TCP port 445. If a vulnerable installation of Microsoft's Windows XP and Windows 2000 is found, the worm utilizes its own FTP server hosted on previously infected machines to download itself onto the newly compromised host. Microsoft's analysis of the worm indicates that it may also spread through port 139. Several variants called Sasser.B, Sasser.C, and Sasser.D appeared within days (with the original named Sasser.A).
Side effects
[edit]An indication of the worm's infection of a given PC is the existence of the files C:\win.log, C:\win2.log or C:\WINDOWS\avserve2.exe on the PC's hard disk, the ftp.exe running randomly and 100% CPU usage, as well as seemingly random crashes with LSA Shell (Export Version) caused by faulty code used in the worm.
The most characteristic symptom of the worm is the shutdown timer that appears due to the worm crashing LSASS.exe.
Mitigation
[edit]The specific hole Sasser exploits is documented by Microsoft in its MS04-011 bulletin (CVE-2003-0533),[3] for which a patch had been released seventeen days earlier.[2] It is easily stopped by a properly configured firewall or by downloading system updates from Windows Update.
Impact
[edit]The impact of Sasser included the news agency Agence France-Presse (AFP) having all its satellite communications blocked for hours and the U.S. flight company Delta Air Lines having to cancel several trans-Atlantic flights because its computer systems had been swamped by the worm. The Nordic insurance company If and their Finnish owners Sampo Bank came to a complete halt and had to close their 130 offices in Finland. The British Coastguard had its electronic mapping service disabled for a few hours, and Goldman Sachs, Deutsche Post, and the European Commission also had issues with the worm. The X-ray department at Lund University Hospital had all their four layer X-ray machines disabled for several hours and had to redirect emergency X-ray patients to a nearby hospital.
Some technology specialists speculated that the worm writer reverse-engineered the patch to discover the vulnerability, which would open millions of computers whose operating system had not been upgraded with the security update.[5]
Author
[edit]On 7 May 2004, an 18-year-old German named Sven Jaschan from Rotenburg, Lower Saxony, then student at a technical college, was arrested for writing the worm. German authorities were led to Jaschan partly because of information obtained in response to a bounty offer by Microsoft of US$250,000.
One of Jaschan's friends had informed Microsoft that his friend had created the worm. He further revealed that not only Sasser, but also Netsky.AC, a variant of the Netsky worm, was his creation. Another variation of Sasser, Sasser.E, was found to be circulating shortly after the arrest. It was the only variation that attempted to remove other worms from the infected computer, much in the way Netsky does.
Jaschan was tried as a minor because the German courts determined that he created the worm before he was 18. The worm itself had been released on his 18th birthday (29 April 2004). Sven Jaschan was found guilty of computer sabotage and illegally altering data. On Friday, 8 July 2005, he received a 21-month suspended sentence.
See also
[edit]Notes
[edit]References
[edit]- ^ Macrae, Duncan (2014-04-11). "Everything you need to know about the Sasser worm". Tech Monitor. Retrieved 2023-02-06.
- ^ a b "Win32/Sasser". Microsoft Security Intelligence. Nov 11, 2004. Archived from the original on 31 October 2022. Retrieved 6 Feb 2023.
- ^ a b "Microsoft Security Bulletin MS04-011 - Critical". learn.microsoft.com.
- ^ "Network Security, Vulnerability Assessment, Intrusion Prevention". 2006-01-09. Archived from the original on 2006-01-09. Retrieved 2023-02-06.
- ^ Net-Worm.Win32.Sasser On a Physical PC Network, 30 April 2014, retrieved 2023-02-06
External links
[edit]- Microsoft Security Bulletin: MS04-011
- CAN-2003-0533
- Bugtraq ID 10108 Archived 2014-05-28 at the Wayback Machine
- Read here how you can protect your PC (Microsoft Security page) - Includes links to the info pages of major anti-virus companies.
- New Windows Worm on the Loose (Slashdot article)
- Report on the effects of the worm from the BBC
- German admits creating Sasser (BBC News)
- Sasser creator avoids jail term (BBC News)
Sasser (computer worm)
View on GrokipediaHistory
Discovery and Initial Spread
The Sasser worm was first detected on April 30, 2004, by antivirus researchers monitoring network traffic for emerging threats.[15] This initial variant, known as Sasser.A, targeted unpatched Windows XP and Windows 2000 systems, exploiting a buffer overflow vulnerability in the Local Security Authority Subsystem Service (LSASS).[16] Within hours of its release, the worm began propagating rapidly across the internet without requiring user interaction, such as email attachments. Starting May 1, 2004, Sasser spread aggressively, infecting hundreds of thousands of computers within the first few days, with early hotspots in the United States, United Kingdom, and parts of Asia.[17] Infected systems exhibited disruptive behaviors, including frequent automatic shutdowns after 60 seconds and blue screen errors, which hampered normal operations and drew widespread attention from users and organizations.[18] Antivirus firms estimated that Sasser and its variants impacted more than one million machines worldwide within days, primarily home users and small networks lacking firewalls or timely patches.[18][3] Over the following week, multiple variants emerged between May 1 and May 6, 2004, adapting the worm's code to evade detection and enhance propagation. Sasser.B appeared on May 1, 2004, improving scanning efficiency for vulnerable hosts through better random IP selection, while Sasser.C emerged on May 4 with further enhancements.[15] Sasser.D surfaced around May 4–7, introducing more aggressive network probing. Later variants introduced additional features: Sasser.B added a backdoor on TCP port 9898; Sasser.C displayed an infection message; Sasser.D added a backdoor on TCP port 1023; Sasser.E, detected around May 8–13 shortly after the creator's arrest, included a backdoor on TCP port 9999 and displayed a warning message urging users to patch their systems via Microsoft's website.[15][19] These rapid iterations contributed to the worm's estimated total infections exceeding one million in its early phase.[18]Development and Attribution
The Sasser worm was developed in April 2004 by Sven Jaschan, an 18-year-old student at a technical college in Rotenburg an der Fulda, Lower Saxony, Germany, as part of his personal experimentation in malware creation.[20] Jaschan, who had limited prior programming experience but was self-taught in virus writing, coded Sasser on a home computer in his bedroom, motivated by a desire to gain recognition as a skilled programmer and potentially assist his mother's job prospects at a security firm.[21] This project followed his earlier work on the Netsky worm family, a series of about 30 variants he authored starting in February 2004, which demonstrated his evolving techniques in worm propagation and anti-antivirus evasion.[22][10] Attribution to Jaschan began shortly after Sasser's emergence on May 1, 2004, when Microsoft and antivirus researchers, including those at F-Secure and Symantec, identified striking code similarities between Sasser and the Netsky family, such as identical random number generation algorithms, shared subroutines for network scanning, and common FTP functions.[22][23] These parallels were reinforced by messages embedded in a Netsky variant (Netsky.AC) explicitly claiming authorship of Sasser, which analysts viewed as a signature from the same developer.[24] Tips from the antivirus community and informal networks further narrowed the search, culminating in two informants—acquaintances of Jaschan—contacting authorities in pursuit of Microsoft's $250,000 bounty for information leading to the worm's creator.[13][25] Jaschan was arrested at his home on May 8, 2004, and confessed to authoring both Sasser and the Netsky variants during initial questioning, amid growing media coverage of the worm's disruptions.[20][26] His admission was corroborated by forensic analysis of his computer, which contained source code matching the worms, solidifying the attribution without prolonged investigation.[27]Technical Details
Exploited Vulnerability
The Sasser worm exploited a critical buffer overflow vulnerability in the Local Security Authority Subsystem Service (LSASS.exe), a core component of the Windows operating system responsible for security policy enforcement and user authentication.[28][29] This flaw, identified as CVE-2003-0533, stemmed from improper handling of Remote Procedure Call (RPC) requests in the LSASRV.DLL library, specifically within Active Directory service functions like DsRolerUpgradeDownlevelServer, which failed to adequately validate input lengths for debug entries in the DCPROMO.LOG file.[29][28] As a result, attackers could trigger a stack-based buffer overflow by sending a specially crafted RPC message, enabling arbitrary code execution with SYSTEM-level privileges without requiring user interaction or authentication.[29][30] The vulnerability was detailed in Microsoft Security Bulletin MS04-011, released on April 13, 2004, which highlighted its potential for remote code execution over TCP port 445 using the Server Message Block (SMB) protocol.[28] Primarily affecting unpatched systems running Windows XP Service Pack 1 (SP1), Windows 2000 Service Packs 2, 3, and 4 (SP2/SP3/SP4), and Windows Server 2003, the flaw rendered these installations susceptible to exploitation from anonymous remote attackers scanning the network.[29][28] Systems with later service packs or those updated via MS04-011 were immune, as the patch addressed the input validation issues in LSASS to prevent overflow conditions.[28] Sasser leveraged this port for initial propagation attempts, though full infection mechanics are covered elsewhere.[30]Propagation Mechanism
The Sasser worm propagates exclusively through network-based mechanisms, targeting vulnerable Windows systems without relying on user interaction such as email attachments or file sharing. Upon execution, it initiates scanning operations by creating 128 concurrent threads that probe random IPv4 addresses for open TCP port 445, which is associated with the Server Message Block (SMB) service. This scanning selects target IP addresses with a bias toward local networks: approximately 50% are fully random, 25% share the first octet with the infected host, and 25% share the first two octets, enabling efficient spread within local environments. The worm can scan over 200 addresses per second per infected host, allowing rapid exploration of potential victims.[31][32] Once a vulnerable system is identified on port 445, Sasser exploits the Local Security Authority Subsystem Service (LSASS) buffer overflow vulnerability (MS04-011) by sending specially crafted, malformed Remote Procedure Call (RPC) packets. This exploit injects shellcode into the target process, granting the worm control without authentication. The shellcode then binds a remote command shell to TCP port 9996 and an FTP server to TCP port 5554 on the infected host, facilitating the download of the worm's executable (such as avserve.exe for the original variant). The target system connects to the attacker's FTP server, retrieves the payload anonymously, executes it, and the cycle repeats as the new instance begins scanning.[31][32] Variants of Sasser, such as Sasser.B, maintain the core propagation strategy but introduce minor modifications, including the use of avserve2.exe as the payload filename and alternative mutex names like "Jobaka3" to prevent multiple infections on the same host. This design avoids mass-mailing or social engineering tactics, allowing the worm to spread silently across local networks and evade early detection by antivirus tools focused on email vectors. The propagation remains confined to direct network exploits, with no reliance on external hosts beyond infected peers for payload distribution.[31][32]Payload Execution
Upon successful exploitation of the stack-based buffer overflow vulnerability in the Local Security Authority Subsystem Service (LSASS), the Sasser worm executes its payload on the infected system. The worm copies its binary to the Windows system directory, typically asavserve.exe for the initial variant, and establishes persistence by adding a registry key under HKLM\SOFTWARE\[Microsoft](/page/Microsoft)\Windows\CurrentVersion\Run to launch the file on startup.[31][32] To avoid multiple concurrent executions, it creates a mutex named "Jobaka3l".[31] The process disguises itself by running under the name lsass.exe in the %System% directory, mimicking the legitimate LSASS service.[31]
The core payload behavior centers on propagation, with the worm spawning up to 128 threads to generate random IP addresses and probe TCP port 445 for vulnerable hosts.[32] Upon identifying a susceptible system, it triggers another buffer overflow to deploy shellcode that instructs the target to connect to the infected machine's FTP server on TCP port 5554, downloading and executing a copy of the worm.[32] Additionally, the payload opens a backdoor listener on TCP port 9996, enabling remote shell access for command execution.[32] The worm lacks capabilities for data exfiltration, encryption, or other destructive actions beyond facilitating its spread.[1]
Variants of Sasser exhibit minor differences in payload execution while retaining the fundamental propagation and backdoor mechanisms. Sasser.A, the original variant, integrates code that leads to frequent system shutdowns approximately every minute due to repeated LSASS crashes induced by the worm's activity.[30] In Sasser.B, the file is named avserve2.exe with an alternate mutex "Jobaka3", and it creates additional temporary files like _up.exe in the system32 directory.[32] Sasser.C, D, and E introduce variations such as increased thread counts (e.g., 1024 in C), different file names (e.g., SKYNETAVE.EXE in D), and enhanced backdoor features for remote control in E, but all prioritize scanning and infection over other payloads.[31]