Likes Likes:  0
Resultaten 1 tot 14 van de 14
Geen
  1. #1
    Larry Seltzer
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Hysterical first technical alert from US-CERT

    I just got the alert below from US-CERT. It's one of the new lists they =
    started. Some
    things about it bother me.

    First, it's dated 1/28, the day MyDoom.B was discovered, and the message =
    sent field says
    that too; other dates in the headers disagree.=20

    Second, and more to the point, it takes an extreme view of MyDoom.B that =
    nobody else is
    supporting, including the sources they cite. MyDoom.B is a flop.

    Am I misreading something? Did anyone else get this on 1/28?

    Larry Seltzer
    eWEEK.com Security Center Editor
    http://security.eweek.com/
    larryseltzer@ziffdavis.com=20

    -----Original Message-----
    From: CERT Advisory [mailto:cert-advisory@cert.org]=20
    Sent: Wednesday, January 28, 2004 7:12 PM
    To: US-CERT Community:
    Subject: US CERT Technical Alert TA04-028A MyDoom.B Rapidly Spreading




    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1


    MyDoom.B Rapidly Spreading

    Mydoom.B is a new variant of the Mydoom worm and is about 29,184
    bytes. This variant attempts to perform a Distributed Denial of
    Service (DDoS) attack against Microsoft.com. Details regarding this
    new worm are still emerging, but it has been validated as spreading =
    in
    the wild. Facts about the worm will be further qualified with follow
    up reports following this initial analysis.

    For the latest information about this worm from US-CERT, readers are
    encouraged to visit =
    http://www.us-cert.gov/cas/techalerts/TA04-028A.html.

    E-mails sent out by Mydoom.B are highly randomized. The From address
    may be spoofed to include one of the following domains: aol.com,
    msn.com, yahoo.com and hotmail.com. A randomized string value may =
    then
    be combined with these to generate new e-mails. This may result in
    overload e-mail servers with many false addresses and auto-replies
    associated with such traffic.

    The subject is randomized to include one of the following
    following:

    * Delivery Error
    * hello
    * Error
    * Mail Delivery System
    * Mail Transaction Failed
    * Returned mail
    * Server Report
    * Status
    * Unable to deliver the message

    The subject may also contain randomized data as seen in a recent live
    sample: "RE: I still love you fLctv".

    The message body is also randomized to include one of the
    following:

    * RANDOMIZED CHARACTERS
    * test
    * The message cannot be represented in 7-bit ASCII encoding and has
    been sent as a binary attachment.
    * sendmail daemon reported: Error #804 occured during SMTP session.
    Partial message has been received.
    * The message contains Unicode characters and has been sent as a
    binary attachment.
    * The message contains MIME-encoded graphics and has been sent as a
    binary attachment.
    * Mail transaction failed. Partial message is available.

    The attachments have a randomized filename selected from one of the
    following string values:

    * body
    * doc
    * text
    * document
    * data
    * file
    * readme
    * message

    The randomized string value is then combined with a randomized
    extension: .exe, .bat, .scr, .cmd or .pif. If the malicious =
    attachment
    is executed, it then opens notepad.exe and displays garbled data
    (binary).

    Once executed, the worm attempts to create the following files in the
    Windows System directory: explorer.exe and dtfmon.dll. The Windows
    registry is then modified to run the worm in memory upon Windows
    startup:

    HKLM\Software\Microsoft\Windows\CurrentVersion\Run
    Explorer=3DC:WINDOWS SYSTEM DIRECTORY\explorer.exe

    The DLL component is associated with a backdoor feature of this worm.
    It is likely that this Trojan worms like the one in Mydoom.A. It =
    scans
    through a range of TCP addresses looking for inbound TCP traffic.
    Inbound TCP traffic can be used to configure the infected computer as
    a proxy computer or to install code of choice on the infected
    computer. More importantly, attackers are already working on tools to
    hijack Mydoom infected computers to install code of choice.

    The DDoS attack of Mydoom.B is against www.microsoft.com. There is
    information claiming that it may also be directed at sco.com, but =
    this
    is unsubstantiated at this time. It appears that the more credible
    data is that it only performs a DDoS attack against =
    www.microsoft.com,
    though a previosu version of the virus is confirmed to attack SCO.

    To spread over the KaZaA P2P network, Mydoom.B creates copies of
    itself in the KaZaA shared directory with randomized filenames.
    Filenames include:

    * attackXP-1.26
    * BlackIce_Firewall_Enterpriseactivation_crack
    * MS04-01_hotfix
    * NessusScan_pro
    * icq2004-final
    * winamp5
    * xsharez_scanner
    * zapSetup_40_148

    A randomized extension is then added to the filename selected above,
    being .exe, .scr, .pif or .bat.

    Mydoom.B attempts to harvest e-mails from Temporary Internet files as
    well as via randomized e-mails aforementioned. It does not include =
    any
    e-mails containing the following strings: abuse, accoun, certific,
    listserv, ntivi, icrosoft, admin, page, the.bat, gold-certs, feste,
    submit, help, service, privacy, somebody, soft, contact, site, =
    rating,
    bugs, your, someone, anyone, nothing, nobody, noone, webmaster,
    postmaster, support, samples, info, root, ruslis, nodomai, mydomai,
    example, inpris, borlan, nai., sopho, foo., .mil, gov., .gov, panda,
    icrosof, syma, kasper, mozilla, utgers.ed, tanford.e, acketst, secur,
    isc.o, isi.e, ripe., arin., sendmail, rfc-ed, ietf, iana, usenet,
    fido, linux, kernel, google, ibm.com, fsf., mit.e, math, unix,
    berkeley and spam.

    Mydoom.B also opens TCP port 10080. The worm contains the following
    string: "sync-1.01; andy; I'm just doing my job, nothing personal,
    sorry".

    Alias: Mydoom, Novarg, Mydoom.B

    Sources:

    F-Secure Corp. (http://www.f-secure.com/v-descs/mydoom_b.shtml),
    Jan. 28, 2004

    Bit Defender
    =
    (http://www.bitdefender.com/bd/site/v...3D1&v_id=3D186)=
    ,=20
    Jan. 28, 2004

    iDEFENSE Intelligence Operations, Jan. 28, 2004 Sensible Security
    Solutions Inc. (http://www.sss.ca/), Jan. 28, 2004

    According to iDEFENSE, this new variant of Mydoom appears to have
    different MIMI data for malicious e-mails. The content type appears =
    to
    be plain text and includes a ZIP extension. Mydoom.A had a content
    type of application/octet-stream and multipart/mixed data. It is
    likely that this newest variant of Mydoom will become very widespread
    in the wild. The first variant had well over 3M interceptions by just
    two sources in the first 18 hours of the outbreak.

    Look for questionable files about 29,184 bytes. Look for notepad.exe
    to be opened, displaying binary data (garbled text). Also look for =
    the
    Windows registry created by the worm.

    Recovery: Remove all files and the Windows registry key modifications
    associated with this malicious code threat. Restore corrupted or
    damaged files with clean backup copies.

    Workaround: Configure e-mail servers and workstations to block file
    types commonly used by malicious code to spread to other computers.
    Block ZIP and executable extensions on the gateway and groupware
    level. Also monitor traffic on the network and block ports associated
    with Mydoom, especially inbound TCP ports for the backdoor Trojan
    component and the outbound TCP 10080 port data. Administrators may
    also find value in monitoring traffic associated with the DDoS
    component. Carefully manage all new files, scanning them with updated
    anti-virus software using heuristics prior to use.

    Vendor Fix: Anti-virus vendors will likely release updated signature
    files to protect against this malicious code in the near future. Some
    anti-virus applications may detect this malicious code heuristically.

    Name of Malicious Code: Mydoom.B
    Aliases:
    Mydoom.B
    Mydoom
    Novarg
    Size in Bytes: 29184
    Subjects: RE: I still love you fLctv
    Body: Error 551: We are sorry your UTF-8 encoding is not supported
    by the server, so the text was automatically zipped and attached to
    this message.
    Attachments: message.zip

    This document was developed based on material contributed by =
    iDEFENSE.
    Our thanks for their contribution.

    Last updated January 28, 2004
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.1 (GNU/Linux)

    iD8DBQFAGEufXlvNRxAkFWARAjEOAJ92cfCtcUVX+/6CGoRwGj7mIbxhzQCg0mdJ
    /ip1ThurA7opfYb0JUET2UI=3D
    =3Dj+iB
    -----END PGP SIGNATURE-----



  2. #2
    Valdis.Kletnieks@vt.edu
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    --==_Exmh_1073267199P
    Content-Type: text/plain; charset=us-ascii

    On Wed, 04 Feb 2004 09:41:39 EST, Larry Seltzer said:

    > The advisory specifically says that MyDoom.B is spreading rapidly, and that was never
    > the case. It didn't say that it *could* spread rapidly. Maybe you think misinforming in
    > order to induce caution is a good idea, but I expect nothing but the truth from an
    > agency like this.


    And I posted a heads-up to our local staff about Dumaru a lot quicker than
    I did for MyDoom, because from where *I* was, I saw a *huge* initial spike
    of Dumaru. If I had waited, I would have realized that Dumaru had fizzled.
    On the other hand, if I had waited that long and it took off like MyDoom,
    we'd have been screwed.

    As I said - would you rather they delayed 12 or 18 hours to identify
    *for sure* how fast it was spreading? Read Nick Weaver's work on
    Warhol Worms at http://www.cs.berkeley.edu/~nweaver/warhol.html and then
    ask yourself how much time they should wait and verify before releasing.

    Unless you have *proof* that they already *knew* it was a snoozer when
    they hit send, or you have *specific* recommendations on how they can
    do better, let it slide.

    Or alternatively, what would *YOU* do if your boss at Ziff Davis told you that
    there were cases where your article *had* to be on the web server *within an
    hour* of you getting the first hint of the story, or real damage might happen?
    Oh, and you don't know which stories those are, and which ones you can afford
    to wait 2 or 3 hours and do follow-ups on first. Oh, and Ziff Davis also said
    that if you screwed up and got a fact wrong, you'd hear about it from all your
    readers.

    If you got a lead that a massive DDoS was coming in 90 minutes, what would you
    do?




    --==_Exmh_1073267199P
    Content-Type: application/pgp-signature

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.3 (GNU/Linux)
    Comment: Exmh version 2.5 07/13/2001

    iD8DBQFAISelcC3lWbTT17ARAvDAAJ9twduiSXgDeBkPRxpf3W pijMKOdQCdFaby
    ZK0t+/Y5ZEiD2V/gOIii2Rg=
    =jwSZ
    -----END PGP SIGNATURE-----

    --==_Exmh_1073267199P--

  3. #3
    Valdis.Kletnieks@vt.edu
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    --==_Exmh_1332279563P
    Content-Type: text/plain; charset=us-ascii

    On Tue, 03 Feb 2004 07:11:49 EST, Larry Seltzer <larry@larryseltzer.com> said:

    > First, it's dated 1/28, the day MyDoom.B was discovered, and the message sent

    field says
    > that too; other dates in the headers disagree.


    Oh, like the fact that a lot of mail servers were getting pounded by MyDoom.*A*
    doesn't mean that there could be delays along the line? (Remember to add in the
    timezones - at least some of the boxes are running in GMT not EST5EDT).

    > Second, and more to the point, it takes an extreme view of MyDoom.B that nobody else is
    > supporting, including the sources they cite. MyDoom.B is a flop.


    OK. So let's see. We've got one highly successful virus (MyDoom.A) on the
    loose at the time of writing, another variant that's essentially identical
    except for the target, and no clear indication why this one *shouldn't*
    take off as well.

    Yes, it took an extreme view that nobody is supporting *NOW*. Now isn't
    last Wednesday night, when there wasn't a week's worth of hindsight.

    Yes, it fizzled. Please point us at the information available to the CERT
    guys *at the time* that proves there was *no* way that MyDoom.B could
    possibly ever be a real threat. What would you have the CERT guys do,
    *not* send the advisory just because they aren't 100% sure at the time?

    I suppose you also understand why MyDoom-A was huge and Dumaru-whatever that
    showed up 2 days before was a yawner. Also, note that I got more copies of
    Dumary in the first 2 hours of THAT one than I got *total* of MyDoom-A - so
    based on the first 2 hours from where *I* am, Dumaru was looking like a much
    bigger event.

    > Am I misreading something? Did anyone else get this on 1/28?


    Received: from lists2.securityfocus.com (lists2.securityfocus.com [205.206.231.20]) by outgoing2.securityfocus.com (Postfix) with QMQP id B5ECF8F5D0; Mon, 02 Feb 2004 12:27:56 -0700 (MST)
    Received: (qmail 11614 invoked from network); Thu, 29 Jan 2004 00:11:38 +0000
    Date: Wed, 28 Jan 2004 19:12:09 -0500

    Looks like some delay there. But it was already at SecurityFocus's qmail
    within seconds (the Date: is actually 31 seconds ahead of the Received: once
    you allow for timezones - somebody isn't using NTP

    --==_Exmh_1332279563P
    Content-Type: application/pgp-signature

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.3 (GNU/Linux)
    Comment: Exmh version 2.5 07/13/2001

    iD8DBQFAIQIzcC3lWbTT17ARAomlAJwPGC6OqK56sy2sVMAQ72 qb2AnfzACg7ziy
    qdjyEJW7ywuNuAV99iB8VWw=
    =ZAYc
    -----END PGP SIGNATURE-----

    --==_Exmh_1332279563P--

  4. #4
    Larry Seltzer
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    RE: Hysterical first technical alert from US-CERT

    >>OK. So let's see. We've got one highly successful virus (MyDoom.A) on =
    the loose at
    the time of writing, another variant that's essentially identical except =
    for the target,
    and no clear indication why this one *shouldn't* take off as well.
    >>...
    >>Yes, it took an extreme view that nobody is supporting *NOW*. Now =

    isn't last
    Wednesday night, when there wasn't a week's worth of hindsight.

    The advisory specifically says that MyDoom.B is spreading rapidly, and =
    that was never
    the case. It didn't say that it *could* spread rapidly. Maybe you think =
    misinforming in
    order to induce caution is a good idea, but I expect nothing but the =
    truth from an
    agency like this.

    Larry Seltzer
    eWEEK.com Security Center Editor
    http://security.eweek.com/
    larryseltzer@ziffdavis.com=20



  5. #5
    Eggers, Bill A [LTD]
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    RE: Hysterical first technical alert from US-CERT

    Seeing as how SCO got shut down and Washington Mutual (I tried to get =
    money from them today and couldn't, they indicated the MyDoom.B was the =
    cause of their problems and several other banks), I'd say it was more =
    than just a minor bother and a little less than hysterical. However, =
    they do have some strong language in their alert. I'm just glad the =
    government is putting some real resources behind the effort.=20

    Bill Eggers
    Tech Support Manager
    SprintDSL

    -----Original Message-----
    From: Larry Seltzer [mailto:larry@larryseltzer.com]
    Sent: Tuesday, February 03, 2004 7:12 AM
    To: bugtraq@securityfocus.com
    Subject: Hysterical first technical alert from US-CERT


    I just got the alert below from US-CERT. It's one of the new lists they =
    started. Some
    things about it bother me.

    First, it's dated 1/28, the day MyDoom.B was discovered, and the message =
    sent field says
    that too; other dates in the headers disagree.=20

    Second, and more to the point, it takes an extreme view of MyDoom.B that =
    nobody else is
    supporting, including the sources they cite. MyDoom.B is a flop.

    Am I misreading something? Did anyone else get this on 1/28?

    Larry Seltzer
    eWEEK.com Security Center Editor
    http://security.eweek.com/
    larryseltzer@ziffdavis.com=20

    -----Original Message-----
    From: CERT Advisory [mailto:cert-advisory@cert.org]=20
    Sent: Wednesday, January 28, 2004 7:12 PM
    To: US-CERT Community:
    Subject: US CERT Technical Alert TA04-028A MyDoom.B Rapidly Spreading




    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1


    MyDoom.B Rapidly Spreading

    Mydoom.B is a new variant of the Mydoom worm and is about 29,184
    bytes. This variant attempts to perform a Distributed Denial of
    Service (DDoS) attack against Microsoft.com. Details regarding this
    new worm are still emerging, but it has been validated as spreading =
    in
    the wild. Facts about the worm will be further qualified with follow
    up reports following this initial analysis.

    For the latest information about this worm from US-CERT, readers are
    encouraged to visit =
    http://www.us-cert.gov/cas/techalerts/TA04-028A.html.

    E-mails sent out by Mydoom.B are highly randomized. The From address
    may be spoofed to include one of the following domains: aol.com,
    msn.com, yahoo.com and hotmail.com. A randomized string value may =
    then
    be combined with these to generate new e-mails. This may result in
    overload e-mail servers with many false addresses and auto-replies
    associated with such traffic.

    The subject is randomized to include one of the following
    following:

    * Delivery Error
    * hello
    * Error
    * Mail Delivery System
    * Mail Transaction Failed
    * Returned mail
    * Server Report
    * Status
    * Unable to deliver the message

    The subject may also contain randomized data as seen in a recent live
    sample: "RE: I still love you fLctv".

    The message body is also randomized to include one of the
    following:

    * RANDOMIZED CHARACTERS
    * test
    * The message cannot be represented in 7-bit ASCII encoding and has
    been sent as a binary attachment.
    * sendmail daemon reported: Error #804 occured during SMTP session.
    Partial message has been received.
    * The message contains Unicode characters and has been sent as a
    binary attachment.
    * The message contains MIME-encoded graphics and has been sent as a
    binary attachment.
    * Mail transaction failed. Partial message is available.

    The attachments have a randomized filename selected from one of the
    following string values:

    * body
    * doc
    * text
    * document
    * data
    * file
    * readme
    * message

    The randomized string value is then combined with a randomized
    extension: .exe, .bat, .scr, .cmd or .pif. If the malicious =
    attachment
    is executed, it then opens notepad.exe and displays garbled data
    (binary).

    Once executed, the worm attempts to create the following files in the
    Windows System directory: explorer.exe and dtfmon.dll. The Windows
    registry is then modified to run the worm in memory upon Windows
    startup:

    HKLM\Software\Microsoft\Windows\CurrentVersion\Run
    Explorer=3DC:WINDOWS SYSTEM DIRECTORY\explorer.exe

    The DLL component is associated with a backdoor feature of this worm.
    It is likely that this Trojan worms like the one in Mydoom.A. It =
    scans
    through a range of TCP addresses looking for inbound TCP traffic.
    Inbound TCP traffic can be used to configure the infected computer as
    a proxy computer or to install code of choice on the infected
    computer. More importantly, attackers are already working on tools to
    hijack Mydoom infected computers to install code of choice.

    The DDoS attack of Mydoom.B is against www.microsoft.com. There is
    information claiming that it may also be directed at sco.com, but =
    this
    is unsubstantiated at this time. It appears that the more credible
    data is that it only performs a DDoS attack against =
    www.microsoft.com,
    though a previosu version of the virus is confirmed to attack SCO.

    To spread over the KaZaA P2P network, Mydoom.B creates copies of
    itself in the KaZaA shared directory with randomized filenames.
    Filenames include:

    * attackXP-1.26
    * BlackIce_Firewall_Enterpriseactivation_crack
    * MS04-01_hotfix
    * NessusScan_pro
    * icq2004-final
    * winamp5
    * xsharez_scanner
    * zapSetup_40_148

    A randomized extension is then added to the filename selected above,
    being .exe, .scr, .pif or .bat.

    Mydoom.B attempts to harvest e-mails from Temporary Internet files as
    well as via randomized e-mails aforementioned. It does not include =
    any
    e-mails containing the following strings: abuse, accoun, certific,
    listserv, ntivi, icrosoft, admin, page, the.bat, gold-certs, feste,
    submit, help, service, privacy, somebody, soft, contact, site, =
    rating,
    bugs, your, someone, anyone, nothing, nobody, noone, webmaster,
    postmaster, support, samples, info, root, ruslis, nodomai, mydomai,
    example, inpris, borlan, nai., sopho, foo., .mil, gov., .gov, panda,
    icrosof, syma, kasper, mozilla, utgers.ed, tanford.e, acketst, secur,
    isc.o, isi.e, ripe., arin., sendmail, rfc-ed, ietf, iana, usenet,
    fido, linux, kernel, google, ibm.com, fsf., mit.e, math, unix,
    berkeley and spam.

    Mydoom.B also opens TCP port 10080. The worm contains the following
    string: "sync-1.01; andy; I'm just doing my job, nothing personal,
    sorry".

    Alias: Mydoom, Novarg, Mydoom.B

    Sources:

    F-Secure Corp. (http://www.f-secure.com/v-descs/mydoom_b.shtml),
    Jan. 28, 2004

    Bit Defender
    =
    (http://www.bitdefender.com/bd/site/v...3D1&v_id=3D186)=
    ,=20
    Jan. 28, 2004

    iDEFENSE Intelligence Operations, Jan. 28, 2004 Sensible Security
    Solutions Inc. (http://www.sss.ca/), Jan. 28, 2004

    According to iDEFENSE, this new variant of Mydoom appears to have
    different MIMI data for malicious e-mails. The content type appears =
    to
    be plain text and includes a ZIP extension. Mydoom.A had a content
    type of application/octet-stream and multipart/mixed data. It is
    likely that this newest variant of Mydoom will become very widespread
    in the wild. The first variant had well over 3M interceptions by just
    two sources in the first 18 hours of the outbreak.

    Look for questionable files about 29,184 bytes. Look for notepad.exe
    to be opened, displaying binary data (garbled text). Also look for =
    the
    Windows registry created by the worm.

    Recovery: Remove all files and the Windows registry key modifications
    associated with this malicious code threat. Restore corrupted or
    damaged files with clean backup copies.

    Workaround: Configure e-mail servers and workstations to block file
    types commonly used by malicious code to spread to other computers.
    Block ZIP and executable extensions on the gateway and groupware
    level. Also monitor traffic on the network and block ports associated
    with Mydoom, especially inbound TCP ports for the backdoor Trojan
    component and the outbound TCP 10080 port data. Administrators may
    also find value in monitoring traffic associated with the DDoS
    component. Carefully manage all new files, scanning them with updated
    anti-virus software using heuristics prior to use.

    Vendor Fix: Anti-virus vendors will likely release updated signature
    files to protect against this malicious code in the near future. Some
    anti-virus applications may detect this malicious code heuristically.

    Name of Malicious Code: Mydoom.B
    Aliases:
    Mydoom.B
    Mydoom
    Novarg
    Size in Bytes: 29184
    Subjects: RE: I still love you fLctv
    Body: Error 551: We are sorry your UTF-8 encoding is not supported
    by the server, so the text was automatically zipped and attached to
    this message.
    Attachments: message.zip

    This document was developed based on material contributed by =
    iDEFENSE.
    Our thanks for their contribution.

    Last updated January 28, 2004
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.1 (GNU/Linux)

    iD8DBQFAGEufXlvNRxAkFWARAjEOAJ92cfCtcUVX+/6CGoRwGj7mIbxhzQCg0mdJ
    /ip1ThurA7opfYb0JUET2UI=3D
    =3Dj+iB
    -----END PGP SIGNATURE-----



  6. #6
    Philip Rowlands
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    On Tue, 3 Feb 2004, Larry Seltzer wrote:

    >I just got the alert below from US-CERT. It's one of the new lists they
    >started. Some things about it bother me.
    >
    >First, it's dated 1/28, the day MyDoom.B was discovered, and the
    >message sent field says that too; other dates in the headers disagree.
    >
    >Second, and more to the point, it takes an extreme view of MyDoom.B
    >that nobody else is supporting, including the sources they cite.
    >MyDoom.B is a flop.
    >
    >Am I misreading something? Did anyone else get this on 1/28?


    Yes; everyone subscribed to the [Carnegie Mellon] CERT/CC list. It's
    been swallowed by US-CERT:

    http://www.cert.org/advisories/us-ce...ouncement.html

    however...

    "significant changes to CERT Advisories that will not only maintain the
    accuracy and fairness that are the hallmarks of CERT Advisories, but
    will also provide more information about more topics than before"

    and

    "We will maintain the same high quality control standards, edit content
    for security and privacy, and work to ensure technical accuracy as well
    as timeliness."

    Thank DoHS for that!

  7. #7
    Andreas Marx
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    Hi!

    >I just got the alert below from US-CERT. It's one of the new lists they
    >started. Some things about it bother me. First, it's dated 1/28, the day
    >MyDoom.B was discovered, and the message sent field says that too; other
    >dates in the headers disagree.


    Yes, I got this report as well and I agree with you that MyDoom.B was not
    able to spread widely at all. It looks like a mistake to me -- maybe they
    wanted to warn for MyDoom.A instead of MyDoomB.?

    At the time of the writing, Trend Micro's online scanner HouseCall has
    found and cleaned more than 900.000 MyDoom.A-infected machines:
    <http://www.trendmicro.com/vinfo/viru...DOOM.A&VSect=S>

    However, they only found 7 (seven) MyDoom.B infected PCs world-wide:
    <http://www.trendmicro.com/vinfo/viru...DOOM.B&VSect=S>

    At least one av company has send out a wrong alert for MyDoom.B as well
    (they wanted to update their MyDoom.A advisory, but they mixed-up both
    worms descriptions).

    A German news about this possible false alert can be found here:
    http://www.heise.de/newsticker/meldung/44281

    cheers,
    Andreas

    --
    BSc. Andreas Marx <amarx@gega-it.de>, http://www.av-test.org
    AV-Test GmbH, Klewitzstr. 7, 39112 Magdeburg, Germany
    Phone: +49 (0)391 6075466, Fax: +49 (0)391 6075469


  8. #8
    Andrew Fried
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    I'm a little surprised by some of the critical reactions to the
    US-Cert's issuance of the MyDoom alerts.

    Being in the federal sector, I can tell you that the predecessor to
    US-CERT (FedCIRC) received ongoing criticism from the government
    computer security circles for untimely advisories. FedCIRC was overly
    cautious about validating information before disseminating it. The
    result was that advisories were released so long after the event that
    they proved to be of little benefit to those of us on the front lines
    trying to mitigate problems. The joke used to be that we'd read about a
    problem on Bugtra or NANOG, then a week later see the same information
    from FedCIRC.

    When DHS formed US-CERT, they held meetings around the country with a
    variety of groups, not just federal security types, and the most
    resounding request they got was to release alerts and advisories as soon
    as possible. Many suggested that late breaking advisories be labeled as
    preliminary, but released just the same. To US-CERT's credit, they
    listened to those requests and what we saw with MyDoom was advisories
    being released within hours of the onset of an incident.

    Behind the scenes, US-CERT has established a number of secure channels
    to facilitate information sharing among federal agencies. They've
    established working groups which include private sector membership.
    They're ramping up some new initiatives that will bring much needed
    resources to the government such as labs to analyze malware. In my
    mind, this group is trying to focus on cybersecurity needs with the same
    intensity that NASA did to get to the moon.

    I'm not trying to make any sales pitch here, and want to state that I do
    not work with for DHS or US-CERT (which is part of DHS).

    Andrew Fried
    Senior Special Agent
    Treasury Inspector General for Tax Administration


  9. #9
    Stephen Samuel
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    Valdis.Kletnieks@vt.edu wrote:
    > On Tue, 03 Feb 2004 07:11:49 EST, Larry Seltzer <larry@larryseltzer.com> said:
    >>Second, and more to the point, it takes an extreme view of MyDoom.B that nobody else is
    >>supporting, including the sources they cite. MyDoom.B is a flop.

    >
    > OK. So let's see. We've got one highly successful virus (MyDoom.A) on the
    > loose at the time of writing, another variant that's essentially identical
    > except for the target, and no clear indication why this one *shouldn't*
    > take off as well.


    Most of the time, I'd expect that a copycat virus is going to be
    much smaller than the original, since that trick-space is now
    going to be saturated.If there's nothing different in the
    distribution of the virus, then I'd say they were silly to
    presume it was going to be anywhere near as big as MyDoom A.
    --
    Stephen Samuel +1(604)876-0426 samuel@bcgreen.com
    http://www.bcgreen.com/~samuel/
    Powerful committed communication. Transformation touching
    the jewel within each person and bringing it to light.

  10. #10
    Mary Landesman
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    They're saying MyDoom.B **BECAUSE** of the bogus CERT reporting. Do you
    really think your bank tellers know what the difference is? They don't. So
    they take it at face value. Our local media keeps referring to MyDoom.B as
    well. Because they got the same "Alert" everyone else in the media got -
    saying MyDoom.B was rapidly spreading.

    CERT did a disfavor to the media, the security company, and the industry
    when they let that shoddy reporting go out on the wire.

    -- Mary


    ----- Original Message -----
    From: "Eggers, Bill A [LTD]" <William.A.Eggers@mail.sprint.com>
    To: <bugtraq@securityfocus.com>
    Sent: Tuesday, February 03, 2004 4:00 PM
    Subject: RE: Hysterical first technical alert from US-CERT


    Seeing as how SCO got shut down and Washington Mutual (I tried to get money
    from them today and couldn't, they indicated the MyDoom.B was the cause of
    their problems and several other banks), I'd say it was more than just a
    minor bother and a little less than hysterical. However, they do have some
    strong language in their alert. I'm just glad the government is putting some
    real resources behind the effort.

    Bill Eggers
    Tech Support Manager
    SprintDSL

    -----Original Message-----
    From: Larry Seltzer [mailto:larry@larryseltzer.com]
    Sent: Tuesday, February 03, 2004 7:12 AM
    To: bugtraq@securityfocus.com
    Subject: Hysterical first technical alert from US-CERT


    I just got the alert below from US-CERT. It's one of the new lists they
    started. Some
    things about it bother me.

    First, it's dated 1/28, the day MyDoom.B was discovered, and the message
    sent field says
    that too; other dates in the headers disagree.

    Second, and more to the point, it takes an extreme view of MyDoom.B that
    nobody else is
    supporting, including the sources they cite. MyDoom.B is a flop.

    Am I misreading something? Did anyone else get this on 1/28?

    Larry Seltzer
    eWEEK.com Security Center Editor
    http://security.eweek.com/
    larryseltzer@ziffdavis.com

    -----Original Message-----
    From: CERT Advisory [mailto:cert-advisory@cert.org]
    Sent: Wednesday, January 28, 2004 7:12 PM
    To: US-CERT Community:
    Subject: US CERT Technical Alert TA04-028A MyDoom.B Rapidly Spreading




    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1


    MyDoom.B Rapidly Spreading

    Mydoom.B is a new variant of the Mydoom worm and is about 29,184
    bytes. This variant attempts to perform a Distributed Denial of
    Service (DDoS) attack against Microsoft.com. Details regarding this
    new worm are still emerging, but it has been validated as spreading in
    the wild. Facts about the worm will be further qualified with follow
    up reports following this initial analysis.

    For the latest information about this worm from US-CERT, readers are
    encouraged to visit http://www.us-cert.gov/cas/techalerts/TA04-028A.html.

    E-mails sent out by Mydoom.B are highly randomized. The From address
    may be spoofed to include one of the following domains: aol.com,
    msn.com, yahoo.com and hotmail.com. A randomized string value may then
    be combined with these to generate new e-mails. This may result in
    overload e-mail servers with many false addresses and auto-replies
    associated with such traffic.

    The subject is randomized to include one of the following
    following:

    * Delivery Error
    * hello
    * Error
    * Mail Delivery System
    * Mail Transaction Failed
    * Returned mail
    * Server Report
    * Status
    * Unable to deliver the message

    The subject may also contain randomized data as seen in a recent live
    sample: "RE: I still love you fLctv".

    The message body is also randomized to include one of the
    following:

    * RANDOMIZED CHARACTERS
    * test
    * The message cannot be represented in 7-bit ASCII encoding and has
    been sent as a binary attachment.
    * sendmail daemon reported: Error #804 occured during SMTP session.
    Partial message has been received.
    * The message contains Unicode characters and has been sent as a
    binary attachment.
    * The message contains MIME-encoded graphics and has been sent as a
    binary attachment.
    * Mail transaction failed. Partial message is available.

    The attachments have a randomized filename selected from one of the
    following string values:

    * body
    * doc
    * text
    * document
    * data
    * file
    * readme
    * message

    The randomized string value is then combined with a randomized
    extension: .exe, .bat, .scr, .cmd or .pif. If the malicious attachment
    is executed, it then opens notepad.exe and displays garbled data
    (binary).

    Once executed, the worm attempts to create the following files in the
    Windows System directory: explorer.exe and dtfmon.dll. The Windows
    registry is then modified to run the worm in memory upon Windows
    startup:

    HKLM\Software\Microsoft\Windows\CurrentVersion\Run
    Explorer=C:WINDOWS SYSTEM DIRECTORY\explorer.exe

    The DLL component is associated with a backdoor feature of this worm.
    It is likely that this Trojan worms like the one in Mydoom.A. It scans
    through a range of TCP addresses looking for inbound TCP traffic.
    Inbound TCP traffic can be used to configure the infected computer as
    a proxy computer or to install code of choice on the infected
    computer. More importantly, attackers are already working on tools to
    hijack Mydoom infected computers to install code of choice.

    The DDoS attack of Mydoom.B is against www.microsoft.com. There is
    information claiming that it may also be directed at sco.com, but this
    is unsubstantiated at this time. It appears that the more credible
    data is that it only performs a DDoS attack against www.microsoft.com,
    though a previosu version of the virus is confirmed to attack SCO.

    To spread over the KaZaA P2P network, Mydoom.B creates copies of
    itself in the KaZaA shared directory with randomized filenames.
    Filenames include:

    * attackXP-1.26
    * BlackIce_Firewall_Enterpriseactivation_crack
    * MS04-01_hotfix
    * NessusScan_pro
    * icq2004-final
    * winamp5
    * xsharez_scanner
    * zapSetup_40_148

    A randomized extension is then added to the filename selected above,
    being .exe, .scr, .pif or .bat.

    Mydoom.B attempts to harvest e-mails from Temporary Internet files as
    well as via randomized e-mails aforementioned. It does not include any
    e-mails containing the following strings: abuse, accoun, certific,
    listserv, ntivi, icrosoft, admin, page, the.bat, gold-certs, feste,
    submit, help, service, privacy, somebody, soft, contact, site, rating,
    bugs, your, someone, anyone, nothing, nobody, noone, webmaster,
    postmaster, support, samples, info, root, ruslis, nodomai, mydomai,
    example, inpris, borlan, nai., sopho, foo., .mil, gov., .gov, panda,
    icrosof, syma, kasper, mozilla, utgers.ed, tanford.e, acketst, secur,
    isc.o, isi.e, ripe., arin., sendmail, rfc-ed, ietf, iana, usenet,
    fido, linux, kernel, google, ibm.com, fsf., mit.e, math, unix,
    berkeley and spam.

    Mydoom.B also opens TCP port 10080. The worm contains the following
    string: "sync-1.01; andy; I'm just doing my job, nothing personal,
    sorry".

    Alias: Mydoom, Novarg, Mydoom.B

    Sources:

    F-Secure Corp. (http://www.f-secure.com/v-descs/mydoom_b.shtml),
    Jan. 28, 2004

    Bit Defender
    (http://www.bitdefender.com/bd/site/v..._id=1&v_id=186),
    Jan. 28, 2004

    iDEFENSE Intelligence Operations, Jan. 28, 2004 Sensible Security
    Solutions Inc. (http://www.sss.ca/), Jan. 28, 2004

    According to iDEFENSE, this new variant of Mydoom appears to have
    different MIMI data for malicious e-mails. The content type appears to
    be plain text and includes a ZIP extension. Mydoom.A had a content
    type of application/octet-stream and multipart/mixed data. It is
    likely that this newest variant of Mydoom will become very widespread
    in the wild. The first variant had well over 3M interceptions by just
    two sources in the first 18 hours of the outbreak.

    Look for questionable files about 29,184 bytes. Look for notepad.exe
    to be opened, displaying binary data (garbled text). Also look for the
    Windows registry created by the worm.

    Recovery: Remove all files and the Windows registry key modifications
    associated with this malicious code threat. Restore corrupted or
    damaged files with clean backup copies.

    Workaround: Configure e-mail servers and workstations to block file
    types commonly used by malicious code to spread to other computers.
    Block ZIP and executable extensions on the gateway and groupware
    level. Also monitor traffic on the network and block ports associated
    with Mydoom, especially inbound TCP ports for the backdoor Trojan
    component and the outbound TCP 10080 port data. Administrators may
    also find value in monitoring traffic associated with the DDoS
    component. Carefully manage all new files, scanning them with updated
    anti-virus software using heuristics prior to use.

    Vendor Fix: Anti-virus vendors will likely release updated signature
    files to protect against this malicious code in the near future. Some
    anti-virus applications may detect this malicious code heuristically.

    Name of Malicious Code: Mydoom.B
    Aliases:
    Mydoom.B
    Mydoom
    Novarg
    Size in Bytes: 29184
    Subjects: RE: I still love you fLctv
    Body: Error 551: We are sorry your UTF-8 encoding is not supported
    by the server, so the text was automatically zipped and attached to
    this message.
    Attachments: message.zip

    This document was developed based on material contributed by iDEFENSE.
    Our thanks for their contribution.

    Last updated January 28, 2004
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.1 (GNU/Linux)

    iD8DBQFAGEufXlvNRxAkFWARAjEOAJ92cfCtcUVX+/6CGoRwGj7mIbxhzQCg0mdJ
    /ip1ThurA7opfYb0JUET2UI=
    =j+iB
    -----END PGP SIGNATURE-----




  11. #11
    Darragh Bailey
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    RE: Hysterical first technical alert from US-CERT

    Quoting "Eggers, Bill A [LTD]" <William.A.Eggers@mail.sprint.com>:

    > Seeing as how SCO got shut down and Washington Mutual (I tried to get money
    > from them today and couldn't, they indicated the MyDoom.B was the cause of
    > their problems and several other banks), I'd say it was more than just a
    > minor bother and a little less than hysterical. However, they do have some
    > strong language in their alert. I'm just glad the government is putting some
    > real resources behind the effort.
    >
    > Bill Eggers
    > Tech Support Manager
    > SprintDSL
    >


    Just 1 minor quibble.

    http://www.sco.com website didn't actually get taken down with the DoS. SCO
    removed the A records so that there is no dns record to use to target the
    website. So nice to see the worm foiled so easily. btw http://sco.com still
    works and points to their website, and all of there sub systems seem to be
    still up, so they effectively rendered the DoS null and void. All the mail
    traffic still being generated is a pain though.

    --
    Darragh

    "Nothing's foolproof to a sufficently talented fool"

  12. #12
    Valdis.Kletnieks@vt.edu
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    --==_Exmh_-1723342094P
    Content-Type: text/plain; charset=us-ascii

    On Thu, 05 Feb 2004 00:33:47 PST, Stephen Samuel said:

    > Most of the time, I'd expect that a copycat virus is going to be
    > much smaller than the original, since that trick-space is now
    > going to be saturated.If there's nothing different in the
    > distribution of the virus, then I'd say they were silly to
    > presume it was going to be anywhere near as big as MyDoom A.


    In general, yes. However, in this case it was a copycat that
    got loose *before* the general population got the meme for "this is
    bad, don't click on it". And remember - they need to get a new
    meme for each specific instance, as the average user can't seem to
    generalize this. (Hmm.. didn't Stephenson's "Snow Crash" talk about
    that?



    --==_Exmh_-1723342094P
    Content-Type: application/pgp-signature

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.3 (GNU/Linux)
    Comment: Exmh version 2.5 07/13/2001

    iD8DBQFAJBAKcC3lWbTT17ARAlRWAKCozW6egmO28FSBz4DNnr dwvYHvjwCgpPq6
    cEmc+mhyi9J3W1djYfJsSHg=
    =imkW
    -----END PGP SIGNATURE-----

    --==_Exmh_-1723342094P--

  13. #13
    Shawn McMahon
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    Re: Hysterical first technical alert from US-CERT

    --Qxx1br4bt0+wmkIi
    Content-Type: text/plain; charset=us-ascii
    Content-Disposition: inline
    Content-Transfer-Encoding: quoted-printable

    On Fri, Feb 06, 2004 at 05:07:06PM -0500, Valdis.Kletnieks@vt.edu said:
    >=20
    > In general, yes. However, in this case it was a copycat that
    > got loose *before* the general population got the meme for "this is
    > bad, don't click on it". And remember - they need to get a new


    One thing nobody seems to want to recognize as a possibility here is
    that MyDoom.B didn't take off as much as MyDoom.A precisely BECAUSE of
    the US-CERT email. A lot of members of the press took that pretty
    seriously, and warnings on the evening network news shows that you
    shouldn't open these attachments are bound to have prevented at least
    some infections.


    --=20
    Shawn McMahon | Let every nation know, whether it wishes us well or ill,
    EIV Consulting | that we shall pay any price, bear any burden, meet any
    UNIX and Linux | hardship, support any friend, oppose any foe, to assure
    http://www.eiv.com| the survival and the success of liberty. - JFK

    --Qxx1br4bt0+wmkIi
    Content-Type: application/pgp-signature
    Content-Disposition: inline

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.0 (GNU/Linux)

    iD8DBQFAJjMkEcl9bQ0RMt0RAtb4AKCKtwKarsjlkTYP/aTdY30FQixP+gCfTDLO
    TgCS5ftAdOm0iWJkYyba4co=
    =d+dI
    -----END PGP SIGNATURE-----

    --Qxx1br4bt0+wmkIi--

  14. #14
    Stephen Martin
    Hysterical first technical alert from US-CERT
    Gast
    n/a Berichten
    Berichten zijn liked



    Thread Starter

    RE: Hysterical first technical alert from US-CERT

    > One thing nobody seems to want to recognize as a possibility here is
    > that MyDoom.B didn't take off as much as MyDoom.A precisely BECAUSE of
    > the US-CERT email. A lot of members of the press took that pretty
    > seriously, and warnings on the evening network news shows that you
    > shouldn't open these attachments are bound to have prevented at least
    > some infections.


    I would describe the MyDoom virus as hitting the main stream media several
    days prior [01.27.03] to CERT's email. Most people [non-tech] that I talked
    to, after 01.27ish had already heard of it and were extremely wary of
    opening
    anything.

    Albeit their e-mail was a catapult to further media coverage, their report
    was misrepresentative of the malware...which should be the bottom line.

    Regards,
    Stephen L. Martin

Webhostingtalk.nl

Contact

  • Rokin 113-115
  • 1012 KP, Amsterdam
  • Nederland
  • Contact
© Copyright 2001-2026 Webhostingtalk.nl.
Web Statistics