| View previous topic :: View next topic |
| Author |
Message |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Tue Feb 21, 2006 5:47 pm Post subject: State of malware detections Feb06 |
|
|
Hello people I would like to share with you the results of some software testing i've undertaken as a result of a new line of "thinking" recently
This comes down to the definition of malware and malware detections as of todays date 22/2/06
Well new netizens might easily believe the following logic
Anti Virus kills Virii&worms etc
Anti trojan goes after Trojans quite possibly
and antispyware shoots down spyware/adware
right ?
Well wrong actually because there all beginning to overlap each other at a rapid rate of knots,its been happening for sometime now and as a community we need to except some home truths
I have a pot of '100' assorted specimens of malware that i've harvested from the darkest corners of the web.
Malware that can be defined as different types of the following>>>
trojans,worms,virii,trackwares,adwares,password stealers,foistware etc
Generally stuff you do not want on your computers
Remembering that detecting malware does'nt always equate to removing malware but at least if a software gives you a "head's up" you can the address the issue so the following tests are based on purely detection of a malware file and not stopping/removal.
With that i've selected a some current popular software's(mix of free and subscription based) to download onto my Pc to pit their detection databases against my pot of misery on my 'puter
Free Anti Spywares>>>
SpyBot-Search_&_Destroy =6/100
AdawareSe =14/100
Windows AS Beta* 34=/100
*Same database definitions as WindowsDefender Beta
Subscription Anti Spywares>>>
CounterSpy =38/100
SypSweeper =30/100
SpywareDoctor =37/100
PestPatrol =10/100
Free AntiTrojan>>>
a2 =50/100
Subscription Anti trojan>>>
Ewido =72/100
TrojanHunter =40/100
Free online AV scanner>>>
Kaspersky AV online scanner =75/100
Free AV softwares>>>
Avast =41/100
AVG =47/100
Well the results speak for themselve's as far as detections go and for the benefit of all here is a copy of the Kaspersky scan log
KASPERSKY ON-LINE SCANNER REPORT
Sunday, February 26, 2006 2:54:48 PM
Operating System: Microsoft Windows XP Home Edition, Service Pack 2 (Build 2600)
Kaspersky On-line Scanner version: 5.0.78.0
Kaspersky Anti-Virus database last update: 26/02/2006
Kaspersky Anti-Virus database records: 167890
Scan Settings
Scan using the following antivirus database standard
Scan Archives true
Scan Mail Bases true
Scan Target Folders
C:\100 malwares\
Scan Statistics
Total number of scanned objects 102
Number of viruses found 54
Number of infected objects 75
Number of suspicious objects 0
Duration of the scan process 00:00:18
Infected Object Name Virus Name Last Action
C:\100 malwares\malware\1.wmv Infected: Trojan-Downloader.WMA.Wimad.d skipped
C:\100 malwares\malware\adload.exe Infected: Trojan-Downloader.Win32.Adload.l skipped
C:\100 malwares\malware\Backdoor.exe Infected: Trojan.Win32.Crypt.e skipped
C:\100 malwares\malware\bagle.exe Infected: Email-Worm.Win32.Bagle.ai skipped
C:\100 malwares\malware\csht.exe/complete_set_hacking_tools+manuals/hacking_tools/hvlscan.zip/UHANFO.EXE Infected: Trojan.DOS.ControlDuSockets.a skipped
C:\100 malwares\malware\csht.exe/complete_set_hacking_tools+manuals/hacking_tools/hvlscan.zip Infected: Trojan.DOS.ControlDuSockets.a skipped
C:\100 malwares\malware\csht.exe/complete_set_hacking_tools+manuals/hacking_tools/wingatespoof_hlp.zip/UHANFO.EXE Infected: Trojan.DOS.ControlDuSockets.a skipped
C:\100 malwares\malware\csht.exe/complete_set_hacking_tools+manuals/hacking_tools/wingatespoof_hlp.zip Infected: Trojan.DOS.ControlDuSockets.a skipped
C:\100 malwares\malware\csht.exe/complete_set_hacking_tools+manuals/hacking_tools/Haktek.exe Infected: HackTool.Win32.Haktek.11 skipped
C:\100 malwares\malware\csht.exe ZIP: infected - 5 skipped
C:\100 malwares\malware\cyrpt.exe Infected: Trojan.Win32.Crypt.e skipped
C:\100 malwares\malware\FT_SilentSudokuInstaller.exe/data0002/data0006 Infected: Trojan-Dropper.Win32.VB.kk skipped
C:\100 malwares\malware\FT_SilentSudokuInstaller.exe/data0002 Infected: Trojan-Dropper.Win32.VB.kk skipped
C:\100 malwares\malware\FT_SilentSudokuInstaller.exe NSIS: infected - 2 skipped
C:\100 malwares\malware\gimmygames9.exe Infected: Trojan-Downloader.Win32.VB.ww skipped
C:\100 malwares\malware\HXDEF.EXE Infected: Trojan-Clicker.Win32.Delf.dm skipped
C:\100 malwares\malware\KeyGen.bat Infected: Trojan.BAT.FoldingHome skipped
C:\100 malwares\malware\keylogger3.exe/rinst.exe Infected: Trojan-Spy.Win32.Agent.f skipped
C:\100 malwares\malware\keylogger3.exe RAR: infected - 1 skipped
C:\100 malwares\malware\keylogger4.exe/Habbo Plus v3.0.exe/bpkhk.dll Infected: Trojan-Spy.Win32.Perfloger.w skipped
C:\100 malwares\malware\keylogger4.exe/Habbo Plus v3.0.exe Infected: Trojan-Spy.Win32.Perfloger.w skipped
C:\100 malwares\malware\keylogger4.exe RAR: infected - 2 skipped
C:\100 malwares\malware\kl1.exe Infected: Trojan-Spy.Win32.Small.dg skipped
C:\100 malwares\malware\loadadv728.exe Infected: Trojan-Downloader.Win32.Harnig.bb skipped
C:\100 malwares\malware\mc-110-12-0000118.exe/data0001 Infected: Trojan-Downloader.NSIS.Agent.p skipped
C:\100 malwares\malware\mc-110-12-0000118.exe NSIS: infected - 1 skipped
C:\100 malwares\malware\mqwkl.exe Infected: Trojan-Downloader.Win32.TSUpdate.p skipped
C:\100 malwares\malware\msshed32.exe Infected: Trojan-Downloader.Win32.Delf.ep skipped
C:\100 malwares\malware\netskyC.exe Infected: Email-Worm.Win32.NetSky.c skipped
C:\100 malwares\malware\password.exe/smssys.exe Infected: Backdoor.Win32.ServU-based skipped
C:\100 malwares\malware\password.exe/svcmgr.exe Infected: Backdoor.Win32.Iroffer.q skipped
C:\100 malwares\malware\password.exe RAR: infected - 2 skipped
C:\100 malwares\malware\Patch.exe Infected: Backdoor.Win32.Optix.b skipped
C:\100 malwares\malware\paytime.exe Infected: Trojan.Win32.StartPage.adi skipped
C:\100 malwares\malware\sdbot.exe Infected: Backdoor.Win32.SdBot.gen skipped
C:\100 malwares\malware\searchbar.exe Infected: Trojan-Downloader.Win32.VB.eu skipped
C:\100 malwares\malware\shell386.exe Infected: Trojan-Downloader.Win32.VB.vb skipped
C:\100 malwares\malware\stub.exe Infected: Trojan-Downloader.Win32.Small.asf skipped
C:\100 malwares\malware\tool2.exe Infected: not-virus:Hoax.Win32.Renos.bj skipped
C:\100 malwares\malware\toolbar.exe Infected: Trojan-Downloader.Win32.VB.vz skipped
C:\100 malwares\malware\trojan1.exe Infected: Email-Worm.Win32.VB.an skipped
C:\100 malwares\malware\trojan10.exe/setup.exe Infected: Trojan-Clicker.HTML.Agent.a skipped
C:\100 malwares\malware\trojan10.exe RAR: infected - 1 skipped
C:\100 malwares\malware\trojan11.exe Infected: Trojan.Win32.Zapchast.ad skipped
C:\100 malwares\malware\trojan13.exe/Setup_toolBar.exe Infected: Trojan-Downloader.Win32.IstBar.nj skipped
C:\100 malwares\malware\trojan13.exe RAR: infected - 1 skipped
C:\100 malwares\malware\trojan14.exe/setup.exe/stream Infected: Trojan-Downloader.Win32.IstBar.no skipped
C:\100 malwares\malware\trojan14.exe/setup.exe Infected: Trojan-Downloader.Win32.IstBar.no skipped
C:\100 malwares\malware\trojan14.exe RAR: infected - 2 skipped
C:\100 malwares\malware\trojan15.exe/data0001 Infected: Trojan-Downloader.Win32.IstBar.lu skipped
C:\100 malwares\malware\trojan15.exe/data0003 Infected: Trojan-Downloader.Win32.IstBar.nn skipped
C:\100 malwares\malware\trojan15.exe NSIS: infected - 2 skipped
C:\100 malwares\malware\trojan16.exe Infected: Trojan-Proxy.Win32.Small.ea skipped
C:\100 malwares\malware\trojan18.exe Infected: Trojan-Dropper.Win32.Delf.qf skipped
C:\100 malwares\malware\trojan19.exe Infected: not-virus:Hoax.Win32.Renos.az skipped
C:\100 malwares\malware\trojan20.exe Infected: Trojan-Dropper.Win32.Agent.agv skipped
C:\100 malwares\malware\trojan22.exe Infected: Backdoor.Win32.SdBot.gen skipped
C:\100 malwares\malware\trojan4.exe Infected: Trojan-Downloader.Win32.Tibs.cc skipped
C:\100 malwares\malware\trojan5.exe Infected: Trojan-Downloader.Win32.Tibser.c skipped
C:\100 malwares\malware\trojan6.exe Infected: Trojan.Win32.Agent.bi skipped
C:\100 malwares\malware\trojan7.exe Infected: Trojan.Win32.Agent.bi skipped
C:\100 malwares\malware\trojan9.exe Infected: Trojan.Win32.VB.aad skipped
C:\100 malwares\malware\Video.exe Infected: Trojan-Dropper.Win32.WinAD.h skipped
C:\100 malwares\malware\winsysban8.exe Infected: Trojan-Clicker.Win32.VB.lg skipped
C:\100 malwares\malware\worm1.exe Infected: Email-Worm.Win32.VB.an skipped
C:\100 malwares\malware\worm2.exe Infected: Email-Worm.Win32.VB.an skipped
C:\100 malwares\malware\worm3.exe Infected: Trojan-Dropper.Win32.VB.lu skipped
C:\100 malwares\malware\worm3.vbs Infected: Email-Worm.VBS.Gedza skipped
C:\100 malwares\malware\worm4.exe Infected: P2P-Worm.Win32.VB.dw skipped
C:\100 malwares\malware\worm5.exe Infected: P2P-Worm.Win32.Wupeer.a skipped
C:\100 malwares\malware\worm6.exe Infected: Email-Worm.Win32.NetSky.q skipped
C:\100 malwares\malware\worm7.exe Infected: P2P-Worm.Win32.VB.dh skipped
C:\100 malwares\malware\worm8.exe Infected: P2P-Worm.Win32.Krepper.c skipped
C:\100 malwares\malware\worm9.exe Infected: not-virus:BadJoke.Win32.VB.p skipped
C:\100 malwares\malware\xload.exe Infected: Trojan-Downloader.Win32.VB.wn skipped
Scan process completed.
*This maybe amateur hour and although some results seem duplicated the file size's/*hash No's varied and a scan using jotti revealed differnt malware titling by other major AV player's
Sidenote If anyone would like to have a copy of malware files for testing/research purposes then drop me a PM  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html
Last edited by fcukdat on Tue Feb 28, 2006 11:50 am; edited 7 times in total |
|
| Back to top |
|
 |
trickyricky Warrior
Joined: 14 Dec 2004 Last Visit: 15 Apr 2009 Posts: 192 Location: London, UK
|
Posted: Wed Feb 22, 2006 2:58 am Post subject: |
|
|
Thanks for doing that research and sharing the results with us, fcukdat.
A couple of observations:
Was Windows Defender really one of the best, percentage-wise, or should its 28/30 score really read 28/80?
What proportion of your festering pot of slimeware falls into each sub-category of malware? Eg: trojans 30%, adware 25&, and so on?
I'm not surprised that Kaspersky performed so well as it's a mature and refined product. However, my experience with AVs generally is that many nasties which are not viri per se are readily detected, but their success at removal on non-virus nasties is usually wanting.
I think I'll PM you for the devilish collection and test some apps myself, to add to your collection of results.  |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Wed Feb 22, 2006 10:19 am Post subject: |
|
|
| trickyricky wrote: |
Was Windows Defender really one of the best, percentage-wise, or should its 28/30 score really read 28/80? |
80 edited thanks
| Quote: |
| What proportion of your festering pot of slimeware falls into each sub-category of malware? Eg: trojans 30%, adware 25&, and so on? |
I will catelogue them shortly the trouble being i grepped them all in 3 sessions(2hrs) and apart from Jotti testing each file for positive malware identification i have'nt actually executed most of them.The one problem i can see is different databases disagree on "titling" of some of the files.
This little project started last saterday morning and is ongoing for a hopefully
I will publish the Jotti results for all files at this topic in time>>>
Test results:Best AV...
| Quote: |
| my experience with AVs generally is that many nasties which are not viri per se are readily detected, but their success at removal on non-virus nasties is usually wanting |
Again this is a "malware detection" test and not a removal&purge test.My interest being in who rings the most alarm bells
your PM is in the post and so is yours Joe  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
EASTER Warrior

Joined: 08 Mar 2005 Last Visit: 01 Feb 2007 Posts: 220 Location: Far Moon Of Endor
|
Posted: Wed Feb 22, 2006 12:20 pm Post subject: |
|
|
| Quote: |
| Again this is a "malware detection" test and not a removal&purge test. My interest being in who rings the most alarm bells |
Grand test. This should at least point out which ones are doing their homework and midnight studies as per a 3rd shift schedule.
I understand some in this line of work actively use an automated process too in their research like a URL Crawler that looks for potential drive-bys and hijacks. _________________ *******************
THE FORCE IS VERY STRONG IN THIS FAMILY!
 |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Wed Feb 22, 2006 1:05 pm Post subject: |
|
|
Thank you fcukdat for performing this test.
I can't believe how poorly Spybot Search & Destroy did!
I am very disappointed. I have made donations to their efforts in the past, and yet progress in improving it has remained very slow. I will wait to see real results before I make any more. |
|
| Back to top |
|
 |
Oldfrog Site Admin

Joined: 08 Aug 2004 Last Visit: 09 Feb 2013 Posts: 1161 Location: Hewitt, TX
|
Posted: Wed Feb 22, 2006 2:12 pm Post subject: |
|
|
| Quote: |
| I can't believe how poorly Spybot Search & Destroy did! |
It would be interesting to see if Spybot would fare any better if the threats were actually installed and active on the system. Watching Spybot scans leads me to believe that it scans by threats rather than starting at the root and working through the folder/file tree. |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Wed Feb 22, 2006 2:33 pm Post subject: |
|
|
| Oldfrog wrote: |
| Quote: |
| I can't believe how poorly Spybot Search & Destroy did! |
It would be interesting to see if Spybot would fare any better if the threats were actually installed and active on the system. Watching Spybot scans leads me to believe that it scans by threats rather than starting at the root and working through the folder/file tree. |
Thank you for sharing that oldfrog.
I have been a big believer in and supporter of SPYBOT for years, and it just kills me, to even think it may be ineffective now. SPYBOT saved my butt years ago, and I have been grateful ever since.
Now, I wonder, could what you said apply to ST and the other programs tested by fcukdat? I don't know enough about this to make a judgment about that. |
|
| Back to top |
|
 |
Oldfrog Site Admin

Joined: 08 Aug 2004 Last Visit: 09 Feb 2013 Posts: 1161 Location: Hewitt, TX
|
Posted: Wed Feb 22, 2006 3:07 pm Post subject: |
|
|
| Quote: |
| could what you said apply to ST and the other programs tested by fcukdat? |
Good question. In another topic crawler_team talked about how ST does detections and mentioned static files. Some products, especially adware, always install in the same place in the same way. So, if a scanner/removal product is looking for that product it may look in the normal places and not detect it if it is sitting idle in a different location. I am not saying that this is either good or bad, just commenting on the fact that it may happen. |
|
| Back to top |
|
 |
hornet777 Warrior Guru

Joined: 28 Oct 2005 Last Visit: 20 Oct 2009 Posts: 458
|
Posted: Wed Feb 22, 2006 6:45 pm Post subject: |
|
|
| channi wrote: |
I can't believe how poorly Spybot Search & Destroy did!
I am very disappointed. I have made donations to their efforts in the past, and yet progress in improving it has remained very slow. I will wait to see real results before I make any more. |
I can. I've been saying for months what a piece of crp it is, but all I got was hostility. It takes 3-4 hours to scan, and doesn't fix anything. Why does it need 35% User and GDI resources? In a word: Delphi: he doesn't know (or want to use) C. Only reason I keep it around at all is for the Immunize function.
Since all the flak I took for going aginst the tide of belief, I just keep my mouth shut. Why bother? Truth means... what? Good work, F'dat. Ich bin verteidigte! |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Wed Feb 22, 2006 8:53 pm Post subject: |
|
|
| Oldfrog wrote: |
| Quote: |
| could what you said apply to ST and the other programs tested by fcukdat? |
Good question. In another topic crawler_team talked about how ST does detections and mentioned static files. Some products, especially adware, always install in the same place in the same way. So, if a scanner/removal product is looking for that product it may look in the normal places and not detect it if it is sitting idle in a different location. I am not saying that this is either good or bad, just commenting on the fact that it may happen. |
I can't imagine how an ASW app which relies on static files could really be effective. I imagine such an app would be easily fooled by simple name changes too wouldn't it?
Thank you so much for teaching me oldfrog. |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Wed Feb 22, 2006 9:05 pm Post subject: |
|
|
| hornet777 wrote: |
| channi wrote: |
I can't believe how poorly Spybot Search & Destroy did!
I am very disappointed. I have made donations to their efforts in the past, and yet progress in improving it has remained very slow. I will wait to see real results before I make any more. |
I can. I've been saying for months what a piece of crp it is, but all I got was hostility. It takes 3-4 hours to scan, and doesn't fix anything. Why does it need 35% User and GDI resources? In a word: Delphi: he doesn't know (or want to use) C. Only reason I keep it around at all is for the Immunize function.
Since all the flak I took for going aginst the tide of belief, I just keep my mouth shut. Why bother? Truth means... what? Good work, F'dat. Ich bin verteidigte! |
Please know I am not discounting your frustrtion nor dismising your feelings. I have pulled it out and swore I would never download it again myself in the past.
I was a bit upset when the Tea Timer GUI was not fixed with the last update, but I went to their SN website and downloaded the fix for it, and now it works fine. I am glad to have it running.
Mr. Kola explained weeks ago why SBS&D is stuck with Delphi problems for now, and as well he promises the next update will be impressive. I am not going to give up on them just yet.
From your description of how loooooooooong it take to scan on your box, it sounds like you may need to uninstall and reinstall a new download of Spybot. I had the same problem with it once, and doing that fixed it. |
|
| Back to top |
|
 |
Nick Site Admin

Joined: 27 Feb 2004 Last Visit: 28 Aug 2012 Posts: 3913 Location: California
|
Posted: Wed Feb 22, 2006 11:04 pm Post subject: |
|
|
Something is not right if Spybot takes over 3 hours to scan. I can scan 250 Gigs in less than 30 minutes.
Yes, Spybot does use alot of system resources, but system resources is the number one reason why Win 98 is out of date. _________________ Nick's Security Ticker
 |
|
| Back to top |
|
 |
EASTER Warrior

Joined: 08 Mar 2005 Last Visit: 01 Feb 2007 Posts: 220 Location: Far Moon Of Endor
|
Posted: Thu Feb 23, 2006 1:31 pm Post subject: |
|
|
KAV displays some very impressive! statistics on that run as does EWIDO which personally is my scanner of choice, very pleased with those results.
As concerns beta, Windows Defender looks to be showing some improvements?
Which of the above programs are also compatible with 98/Me systems?
I know some of you might argue differently but with the onset of HIPS programs, i no longer see any reason why Vendors would not consider addressing again those platforms. Just a personal notion of mine since mentioning 98/Me cannot be secured simply cannot in all honesty hold water anymore when running a SSM for one example. _________________ *******************
THE FORCE IS VERY STRONG IN THIS FAMILY!
 |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Thu Feb 23, 2006 1:51 pm Post subject: |
|
|
Well people i've got a cauldron of crud to test the scanners databases against.It now stands at 100 specimens and thats going to be the final test tally.
I'm going to edit my opening post within the next few days to show new/revised tests because i'm going to test additional softwares as well as retesting of original selection.
New software to be tested>>>
Anti-spyware
PestPatrol
Anti-virus
Avast
AVG
Anti-trojan
a2
TrojanHunter
And hopefully at some point(time allowing) i will post screenshots of Jotti's scan reports against the "100" for validation/hash referencing
*if anyone has any other softwares that they would like tested against the "100" drop me a PM/download link  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
Turtledove SWW Graduate

Joined: 26 Dec 2005 Last Visit: 04 Feb 2011 Posts: 343 Location: California
|
Posted: Fri Feb 24, 2006 12:13 am Post subject: |
|
|
Many thanks for your efforts  |
|
| Back to top |
|
 |
wawadave Warrior Obsessed

Joined: 25 Jan 2004 Last Visit: 24 Jul 2009 Posts: 3448 Location: Illegitimus non carborundum
|
|
| Back to top |
|
 |
Oldfrog Site Admin

Joined: 08 Aug 2004 Last Visit: 09 Feb 2013 Posts: 1161 Location: Hewitt, TX
|
Posted: Fri Feb 24, 2006 11:33 am Post subject: |
|
|
| Quote: |
I can't imagine how an ASW app which relies on static files could really be effective. I imagine such an app would be easily fooled by simple name changes too wouldn't it?
Thank you so much for teaching me oldfrog. |
This is sort of an interesting area. There are some threats that always install in one or more standard ways. Suppose, for example, that you were writing signatures for detecting a BHO. For the BHO to actually pose a threat it has to have a registry entry that "tells" IE that is is installed and active. Without that entry there is no threat. Take the following HJT line from a Smitfraud infection:
| Quote: |
| O2 - BHO: (no name) - {3bf1f86f-b1a8-489b-8d8b-43781d51411f} - C:\WINDOWS\system32\hpEA3E.tmp |
In this case the CLSID is non-random but the filename is random. A scanner can go through the registry and when it detects the CLSID can then determine the file name. It can then proceed to remove the file and the registry entry. If the file is present on the system but the entry is not in the registry it will do neither. Does this constitute a failure of the scanner? My contention is that it does not because there was no threat present to begin with. I will readily admit that installers are a different matter.
This seems to me to be the approach taken by Spybot. During a Spybot scan there is no display indicating that x files have been scanned. Regardless of the number of files on the system there is a display at the bottom that shows a total number of threats and indicates progress based on how many have been checked at that point. This is in effect the same approach used by HJT log helpers. Look at the running processes, the browser default pages, the installed BHOs and Toolbars, the startup items, the LSP stack and running services, in short everywhere that installed malware is likely to leave a signature. Analyzing those allows backtracking to files and folders which can then be removed. Sure, you may miss inactive files or those which aren't where they would need to be to execute properly but those don't pose an active threat either. |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Fri Feb 24, 2006 11:53 am Post subject: |
|
|
Oldfrog,
Thank you very much for the information. That was really helpful.
 |
|
| Back to top |
|
 |
Oldfrog Site Admin

Joined: 08 Aug 2004 Last Visit: 09 Feb 2013 Posts: 1161 Location: Hewitt, TX
|
Posted: Fri Feb 24, 2006 12:17 pm Post subject: |
|
|
Well, while I believe that the information presented is valid the conclusions are all mine so feel free to take pot shots at them. This also points out a fundamental difference between virus scanners and AS scanners.
AV products have typically had to detect single files with particular signatures unrelated to location and have done an excellent job over the years. As a real life example, I have a captive LOP installer sitting in a folder that I have called "Vault" and the only installed scanner currently detecting it is NOD32. I know that other products scan it because NOD32 alerts whenever they access it to do so.
AS products, on the other hand, can't target against single files but have to target against threats which typically involve a number of registry entries as well as both folders and files. These can actually run as applications and have a supporting set of files just like all the legitimate applications on the system. This makes for much more complicated signatures and I hestitate to think how long it would take to scan a system if every file were compared to a list of every file ever known to be installed as part of a threat.
This is probably an over simplification but at least illustrates differences in scanning technique. I actual practice there are combinations used and striking that balance undoubtedly has a lot to do with the effectiveness, or not, of any particular product. |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Sun Feb 26, 2006 4:44 am Post subject: |
|
|
Sorry for the delay....
Revised&test results coming soon after a snafu with AVG Auto-heal function taking out a percentage of nasties.Like i openly admit this is "Amateur hour" tests and you live and learn
Now the revised 100 test specimens are backed up on alternative storage.So with that there is still '100' malware files but not all the original 80 intact so results will vary as an outcome to this snafu
They are comfirmed malware by at least one or more of Jotti's databases.I believe the files are executed in a sandbox and scanned and are not statically scanned unlike these tests
Types of malware roughly fall under the following>>
Trojans = 51
Worms = 11
Keylogger = 4
Dialer = 1
Adware = 33
Oldfrog,initial results do pose some questions reguardless of detection methodology used by the test softwares.
All software's are being pitted against the same folder containing inactive malware.If you say that some scanners only detect running malware then there score should = 0 in all theories.If they only detect a small number then the question has to be why only detect some and not all in their database
FWIW if someone has a file containing a SDbot called "setup.exe" on their PC and it has not been executed.Is that file malware,dose it pose a risk to the persons Pc security ? Should it be expunged ?
Initial testing is suggesting that the AV's&AT's are trouncing the botkillers into the dust for static malware recognition
*unable to test a2 until i get my hands on an activation key(ISP related problem )
**Unable to retest WinDefender again due to a "0x80070663" error on attempted reinstall.Don't ya just love Beta software  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Tue Feb 28, 2006 12:06 pm Post subject: |
|
|
Hi all,sorted a work around for my last two problems.Final results are in for this bout of testing
Final test results>>>
a2 registration issue sorted by using there online scanner which is the same database as the software.
WindowsDefender install prooblem was not going to resolve in a hurry so i've downloaded the original MSAS beta and updated to the current definitions which are the same for both softwares  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
EASTER Warrior

Joined: 08 Mar 2005 Last Visit: 01 Feb 2007 Posts: 220 Location: Far Moon Of Endor
|
Posted: Tue Feb 28, 2006 4:03 pm Post subject: |
|
|
Pest Patrol checks in with an enemic 10? out of possible 100? Is the Rogue List still available?
No, seriously, that is one sad result indeed that borders on useless?
I will say this, very surprising results in those detections with some of them.
Still very impressed with the Ewido returns. _________________ *******************
THE FORCE IS VERY STRONG IN THIS FAMILY!
 |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Tue Feb 28, 2006 6:08 pm Post subject: |
|
|
Thank you fcukdat! Great work.
I uninstalled Spybot S&D this evening, and I am going pay for my trial version of Ewido. |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Wed Mar 01, 2006 8:52 am Post subject: |
|
|
| In light of these test results do you think maybe the Trustworthy Anti-Spyware Products section of the rogue list page should be updated? |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Wed Mar 01, 2006 10:19 am Post subject: |
|
|
| channi wrote: |
| In light of these test results do you think maybe the Trustworthy Anti-Spyware Products section of the rogue list page should be updated? |
No,absolutely not.
You must realise the tests were for detection/identification of inactive malware related files ie they were not running at the time of scanning.This might point to some short comings in certain software but fwiw different programmes work on different levels/methodology of scanning some better,some not
I would like at some point to test software against these malwares whilst they go active but that will take a huge amount of time&resources to achieve so thats for a later date hopefully
FWIW virtually all malware behaves like "trojan" in nature and that includes adware installers/softwares.
The eccentric behaviour i've seen in the tests being the AV's flag more adware/adware installers than the Botkillers
PestPatrol for example flags the Netsky virus,gedza worm but miss's virtually all of the trojan installers/adware setup's
Is this rogue/suspect behaviour,not really just a mediocre software inaction IMO _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
Oldfrog Site Admin

Joined: 08 Aug 2004 Last Visit: 09 Feb 2013 Posts: 1161 Location: Hewitt, TX
|
Posted: Wed Mar 01, 2006 10:57 am Post subject: |
|
|
| Quote: |
| The eccentric behaviour i've seen in the tests being the AV's flag more adware/adware installers than the Botkillers |
That actually makes sense. AVs provide protection by examining every file that enters the system or is opened. AS apps typically provide protection by preventing certain changes being made to the system or certain detected files/file types from executing. I can watch NOD32 alerting on installer files as they are being scanned, without being detected, by a number of AS products. NOD32 has so far not alerted when scanning other components installed by those same installers.
This is not to say that fcukdat's results are not valid nor interesting, only that they are only useful in the context of the testing methodology. With a different methodology the results could be far different and have not necessarily more, but at least different, validity than these.
For a very nice discussion of a different methodology see this. This second methodology seems to be the direction that fcukdat wants to move in. He is correct, though, about the investment in time required to do so. There is also the matter of setting up a testing environment that is both safe and not in danger of becoming contaminated between tests. I am not suggesting that this methodology is the "final answer" as there are metrics lacking in it that I would like to see and it neglects the issue of "new" threats. |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Wed Mar 01, 2006 1:29 pm Post subject: |
|
|
| Oldfrog wrote: |
| AVs provide protection by examining every file that enters the system or is opened. |
LOL You can say that again everytime i go bug hunting i keep forgeting to turn off Avast realtime protection
| Quote: |
| For a very nice discussion of a different methodology see this. |
JMHO One of the excellent resources at SWW forums but maybe passing its sell by date and needs bringing into 06 with 06 databases&malware.
But thats one hell of an ask
On a side note OT but purely comparitive
I would have loved to see a set of results of the test malware that could install after BoClean was installed but that would not be comparing like for like and would render any botkiller surplus to requirement  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
Oldfrog Site Admin

Joined: 08 Aug 2004 Last Visit: 09 Feb 2013 Posts: 1161 Location: Hewitt, TX
|
Posted: Wed Mar 01, 2006 1:47 pm Post subject: |
|
|
| Quote: |
| I would have loved to see a set of results of the test malware that could install after BoClean was installed |
Yes, or perhaps DefenseWall or any of the other HIPS products that are becoming available. That, however, is a different test altogether.
I have probably talked enough about how such a simple thing as a firewall prevented me from getting badly bitten by the .wmf exploit before it had even had any advisories issued but it does show that there are many ways to skin the proverbial cat (as opposed to a real one, no cats were harmed in preparing this post!). |
|
| Back to top |
|
 |
channi Warrior
Joined: 23 Dec 2005 Last Visit: 12 May 2006 Posts: 267
|
Posted: Wed Mar 01, 2006 2:00 pm Post subject: |
|
|
Thank you fcukdat for your speedy reply, and thanks Oldfrog for your good info about this too.
Of course it is going to take me a while to digest it, and it will take even longer for me to catch up, if ever I can at all.
So much to learn!
My head hurts as it is now.
. |
|
| Back to top |
|
 |
herbalist Warrior Addict

Joined: 28 Aug 2004 Last Visit: 25 Jun 2008 Posts: 726 Location: northern Michigan
|
Posted: Wed Mar 01, 2006 2:46 pm Post subject: |
|
|
| Quote: |
| Quote: |
| I would have loved to see a set of results of the test malware that could install after BoClean was installed |
Yes, or perhaps DefenseWall or any of the other HIPS products that are becoming available. That, however, is a different test altogether.
|
I've been trying many of these plus others I've been collecting against System Safety Monitor. So far, the only way I've been able to infect my test unit with them is to click "permit" when it prompts me. So far, nothing has got past it when I've said "No".
I can't compare SSM to any of the other brands as they chose not to make theirs compatible with 98, but if they perform as well as SSM has, the anti-spyware and anti-trojan programs could end up obsolete for prevention purposes. They may have a place for cleaning, but the HIPS beat them hands down at keeping things out, as long as you don't click "allow".
Rick |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Wed Mar 01, 2006 2:50 pm Post subject: |
|
|
Hey Channi don't worry,were all on learning curves and get those days where the sponge gets saturated
| oldfrog wrote: |
| Yes, or perhaps DefenseWall or any of the other HIPS products that are becoming available. That, however, is a different test altogether. |
Staying OT but continuing my public learning curve
I thought Boc was definition based antitrojan that used an alternative kill process on malware when a predefined file tried to execute in the memory
I've not looked at Defencewall &HIPS are they not rule based ala ProcessGuard,SSM and use enduser desicion making to grant rules  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
EASTER Warrior

Joined: 08 Mar 2005 Last Visit: 01 Feb 2007 Posts: 220 Location: Far Moon Of Endor
|
Posted: Wed Mar 01, 2006 6:28 pm Post subject: |
|
|
| Quote: |
I can't compare SSM to any of the other brands as they chose not to make theirs compatible with 98, but if they perform as well as SSM has, the anti-spyware and anti-trojan programs could end up obsolete for prevention purposes. They may have a place for cleaning, but the HIPS beat them hands down at keeping things out, as long as you don't click "allow".
Rick |
It would do well to point out that SSM as the only intelligent HIP vendor showcasing how very secure those platforms now can become, many of those who thought it some advantage to them to disadvantage 98 and Me users might wish to think again or else lose out completely in that area. Thanks to all those security softs vendors who never turned their back on those users.
You are definitely a true credit in contributing to overall internet safety and security and honestly deserve added recognition in that regard.
Come to think of it, i don't know of any other vendor/development aside from System Safety Monitor that is including the 98/Me platforms, so a big plus for them eh?
Without a doubt a real turning point is finally introduced in securing windows platforms to a reasonably acceptable level.
With the new introductions of HIPS technology, prevention with REAL security is a new reality that must be taken seriously. Also since those are still few yet, you can count on programs like SSM and others becoming malware target #1 when the reality and proof of all this finally begins to sink in for everyone.
Confidence IMHO is now turned up several notches, AFAIK the attack methods will need now to focus on yet other alternative windows of opportunity to circumvent these defenses against passage let alone trying to cleverly masquarade themselves to drop their payloads.
With all the recent publicity over rootkits which have directly lead to more intensive research and solutions into just those type detections alone, it's spawned developments that at the same time are effectively making null those common malware threats that botkillers have been trying to get a handle on for years with signature-based methods and the occasional "resident" theory offered by most.
In addition and to return again to Topic, Thank You is in order to
fcukdat for making the effort to display these comparisons.
Looking at just these results above, it once again shows proof positive the neccesity that HIPS prevention can offer in the way of a more SOLID security with that extra layer of protection. _________________ *******************
THE FORCE IS VERY STRONG IN THIS FAMILY!
 |
|
| Back to top |
|
 |
aigle Newbie
Joined: 02 Mar 2006 Last Visit: 02 Mar 2006 Posts: 1
|
Posted: Thu Mar 02, 2006 2:26 am Post subject: One point |
|
|
I want to mention one thing, people do expect AV to detect other malware also buy they don,t expect Antispywares to catch typical viruses, that,s gob of AV. So if you want to check antispwares, you should include all other types of malware except typical AV. For me personally Kaspersky online scan results are surprising, I don,t think it can detect spywares more than any typical antispyware product. _________________ Muhammad S Hameed |
|
| Back to top |
|
 |
Mrkvonic Warrior
Joined: 27 Sep 2004 Last Visit: 01 Feb 2007 Posts: 209
|
Posted: Thu Mar 02, 2006 4:19 am Post subject: |
|
|
Hello,
A question fcukdat:
How much overlap is there?
How many threats were covered by let's say Ewido + Ad-Aware + Spybot + MSAS?
Maybe this is difficult to analyze, but it could be interesting.
Mrk |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Thu Mar 02, 2006 11:00 am Post subject: Re: One point |
|
|
| aigle wrote: |
| I want to mention one thing, people do expect AV to detect other malware |
Hi aigle and welcome to the SWW forums
I'm not sure i agree with you its not not a level of expectancy thing.
Over the last 12months AV companies have been dev'ing their databases towards "spyware" because they smell the money that is being generated by this neck of the malware community."Spyware" has become public enemy No1 thanks principly to "Adwares".
Lets be honest Blaster(worm) will make a mess of an unprotected system but it dose'nt steal your data for a third party whether its a unethical company or criminal individual.It just borks your PC
The AV's have been long established(old boys) where as the botkillers are the new kids on the block targeting their own little enclave of the malware sphere that up until recent times the AV's have ignored
The AV's have got the $'s & resources to effectively tackle "spyware" as well as worms/Virii etc
Where as the same cannot be said about the botkillers  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
Oldfrog Site Admin

Joined: 08 Aug 2004 Last Visit: 09 Feb 2013 Posts: 1161 Location: Hewitt, TX
|
Posted: Thu Mar 02, 2006 11:14 am Post subject: |
|
|
| I think that an additional factor is the fact that a malware installer, like a virus, is a single file that will normally have a well defined signature. A far different thing than an installed malware product with registry keys, folders/files, etc. Dectection of the installer is far simpler and if done immediately at the point of entrance renders the rest unnecessary. Since my AV is already checking the entry points and file openings for viruses it seems appropriate to me that it also check for installer signatures. I sure don't need yet another product with the same hooks watching the same events (yes, Rick, I am exempting rule-based HIPS from the discussion). I also don't want my AV definitions cluttered with the entire file trees installed by every species of malware. |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Thu Mar 02, 2006 11:17 am Post subject: |
|
|
| Mrkvonic wrote: |
Hello,
A question fcukdat:
How much overlap is there?
How many threats were covered by let's say Ewido + Ad-Aware + Spybot + MSAS?
Maybe this is difficult to analyze, but it could be interesting.
Mrk |
Lots of overlap,the AV's&AT's detect more "Spyware/Adware" the the botkileers detct worms/virii
Thats some serious data greping to be done.Give me a while and i will see what i can do
Actually in hindsight there would not be much point to this info because initial inspection of data concludes that even with all scanners results combined,only 96 specimens were detected
Your chosen combo scored=78/100 identified.
Fwiw one of the Vendors that contacted me for a copy of the "100" now detects all 100 in static mode wherever i place the folder on my HD.But all they have done is added match hash* No's to their scanner
Is this a good thing,well i think so to some respect but fwiw there maybe another 100K+ malware files out there.
How will there scanner perform against my next "100" when they are harvested or the rest of the malware out there
So one fact that is standing out a mile for me is no individual/combo of malware definition based software are up to the task of detecting all malware.Add to that 0day and in the wild malware it opens up big holes in your Pc defence's.
Afterall definition based software are always playing catchup
I've got 2 other projects underway now and they will take some time to finish.The first is a test of realtime detection of malware as it goes active and secondly post infection cleaning of an infected enviroment.
Unfortunetly due to the amount of time it will take to do these two projects the number of malware specimens will be reduced down and so to the number of softwares to be tested
But inorder to stay inline with the current tests the test specimens will be a couple of representitives from each type of malware vs representitives of each kind of antimalware software
FWIW i will also publish the results of the realtime stopping ability of my two favoured rule based protection softwares(Winpatrol&ProcessGuard) vs the test malwares  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
herbalist Warrior Addict

Joined: 28 Aug 2004 Last Visit: 25 Jun 2008 Posts: 726 Location: northern Michigan
|
Posted: Thu Mar 02, 2006 1:40 pm Post subject: |
|
|
| Quote: |
| I also don't want my AV definitions cluttered with the entire file trees installed by every species of malware. |
Most of them are getting to that point anyway. I used to keep a copy of F-Prot for DOS on a 3 floppy set up until last year. It was very handy on 98 and ME units. The signature files have gotten so big that it can't be used this way anymore. Almost 6mb worth now. Same for AntiVir. It wouldn't suprise me that this is part of the reason AVs are getting more demanding on the system.
Rick |
|
| Back to top |
|
 |
hornet777 Warrior Guru

Joined: 28 Oct 2005 Last Visit: 20 Oct 2009 Posts: 458
|
Posted: Thu Mar 02, 2006 3:38 pm Post subject: |
|
|
| Wow, me too, but now (with the big file sizes) I just make a multi-volume ZIP archive, and carry along a copy of PKZIP (DOS) in case I need it. Most of the rest of the stuff (other than boot disk) I just burn onto a CD these days. |
|
| Back to top |
|
 |
fcukdat Warrior Addict

Joined: 01 Jan 2005 Last Visit: 08 Apr 2009 Posts: 757 Location: Yeovil,England.
|
Posted: Thu Mar 02, 2006 3:48 pm Post subject: |
|
|
well it just gose to show the bigger the database,oh er the bigger the update file....
Hang on,the more features,types of scanning surely equals more bloat and resource hogging
In that case effective¤t definitions based software will continue to grow with the rate of the increase in malware  _________________ Malware hunter....Got Bot ?
MIRT Handler >>>
http://www.castlecops.com/c55-MIRT.html |
|
| Back to top |
|
 |
|