home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.vms
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!moe.ksu.ksu.edu!ux1.cso.uiuc.edu!news.cso.uiuc.edu!jsue
- From: jsue@ncsa.uiuc.edu (Jeffrey L. Sue)
- Subject: Re: HELP!!! Security problem for gurus.
- References: <B1FB21FFA27F004AEF@imimnvx.irfmn.mnegri.it> <Bz1nrE.ALq@unx.sas.com> <1992Dec19.025940.1@us.oracle.com>
- Message-ID: <1992Dec22.161918.9033@ncsa.uiuc.edu>
- Originator: jsue@pluto.ncsa.uiuc.edu
- Sender: usenet@news.cso.uiuc.edu (Net Noise owner)
- Organization: The Dow Chemical Company
- Date: Tue, 22 Dec 1992 16:19:18 GMT
- Keywords: hack security files-11 protection
- Lines: 163
-
- In article <1992Dec19.025940.1@us.oracle.com> comet@us.oracle.com writes:
- >In article <Bz1nrE.ALq@unx.sas.com>, sasjzs@falcon.unx.sas.com (Joseph Slater) writes:
- >>
- >>In article <B1FB21FFA27F004AEF@imimnvx.irfmn.mnegri.it>,
- >>PSI%ITAPAC.22800002::PITCLS::ADRIANO@imimnvx.irfmn.mnegri.it
- >>(Adriano Santoni) writes:
- >>
- >>|> I need to avoid certain people to scan a directory of mine.
- >
- >Well, there are "avoidance" techniques you can use.
- >
- >>|> This could seem a very trivial issue, if it was not for
- >>|> the following:
- >>|> o Some users of mine have (and need to retain)
- >>|> *ALL* privileges
- >
- >There are NO "prevention" techniques effective against fully privileged users.
- >
- >Having the directory level be at level 9 or lower (to 15?) and accessing it
- >via a concealed, terminal logical name is an excellent avoidance technique.
- >I've used this in conjunction with recursive directory specifications
- >(by $ CREATE/DIRE <.1> and $ SET FILE/ENTER=[.1]2.DIR 1.DIR. Horrid, eh? ;)
- >to really confuse things. Again, making a pathological directory will stick
- >out during an $ ANALYZE/DISK operation.
- >[some deleted]
- >Other ideas: You can write a program to set the directory (and enclosed files)
- >de-access locked. I don't know how effective (if at all) this would be, but
- >you can $ UNLOCK file when you're ready to use it.
- >
- >You can munge the .DIR file so that anybody normally browsing your directory
- >will receive any escape sequence you can imagine that would disable their
- >terminal. This may discourage casual scanning, and draw the attention of
- >"hacker" types, so you can use this technique both as a decoy and as a
- >countermeasure. I put some of these in higher levels of my recursive tree.
-
- I want to warn anyone using "unsupported" VMS techniques like putting the
- directory below 8 levels deep, renaming the .DIR to something else (e.g.,
- .DAT), or completely munging the directory.
-
- If this information is important to you, then just realize that VMS BACKUP
- will "lose" this information in the event of a full disk restore. Also,
- if ANALYZE/DISK/REPAIR (before or after the restore) will place these files
- in [SYSLOST].
-
- Just to see what would happen, I used my V5.5-2 system to create a directory
- 16 levels deep and created a file in it. This is [X1...X16]XXXX.XXXX
-
- Also, I created [YY]TEST1.DAT and then renamed yy.dir to ZZ.DAT. Here is
- the backup listing that resulted.
-
- *********************
- Listing of save set(s)
-
- Save set: X.BAK
- Written by: S084349
- UIC: [000001,000005]
- Date: 22-DEC-1992 11:02:20.98
- Command: BACKUP/IMAGE DISK$VUSER2: DISK$VUSER1:[000000]X.BAK/SAVE
- Operating system: VAX/VMS version V5.5
- BACKUP version: V5.5-2
- CPU ID register: 13000202
- Node name: _CRVS02::
- Written on: _$1$DKA200:
- Block size: 32256
- Group size: 10
- Buffer count: 116
-
- Image save of volume set
- Number of volumes: 1
-
- Volume attributes
- Structure level: 2
- Label: VUSER2
- Owner:
- Owner UIC: [000001,000004]
- Creation date: 9-DEC-1992 07:39:39.87
- Total blocks: 1316751
- Access count: 3
- Cluster size: 3
- Data check: No Read, No Write
- Extension size: 5
- File protection: System:RWED, Owner:RWED, Group:RE, World:
- Maximum files: 164593
- Volume protection: System:RWCD, Owner:RWCD, Group:RWCD, World:RWCD
- Windows: 16
- Minimum retention: 30 00:00:00.00
- Maximum retention: 30 00:00:00.00
-
- [000000]BACKUP.SYS;1 0 9-DEC-1992 07:39
- [000000]BADBLK.SYS;1 0 9-DEC-1992 07:39
- [000000]BADLOG.SYS;1 0 9-DEC-1992 07:39
- [000000]BITMAP.SYS;1 109 9-DEC-1992 07:39
- [000000]CONTIN.SYS;1 0 9-DEC-1992 07:39
- [000000]CORIMG.SYS;1 0 9-DEC-1992 07:39
- [000000]DEFRAG.DIR;1 1 12-DEC-1992 14:36
- [DEFRAG]$1$DKA0_PIC.LIS;1 13 12-DEC-1992 15:33
- [DEFRAG]$1$DKA200_PIC.LIS;1 10 12-DEC-1992 16:56
- [000000]INDEXF.SYS;1 1062 9-DEC-1992 07:39
- [000000]OFFLOAD.DIR;1 1 12-DEC-1992 14:41
- [000000]PATCHES.DIR;1 1 10-DEC-1992 09:08
- [000000]SCR.DIR;1 1 12-DEC-1992 14:41
- [000000]SYSLOST.DIR;1 1 9-DEC-1992 07:39
- [000000]VOLSET.SYS;1 0 9-DEC-1992 07:39
- [000000]X1.DIR;1 1 22-DEC-1992 10:48
- [X1]X2.DIR;1 1 22-DEC-1992 10:48
- [X1.X2]X3.DIR;1 1 22-DEC-1992 10:48
- [X1.X2.X3]X4.DIR;1 1 22-DEC-1992 10:48
- [X1.X2.X3.X4]X5.DIR;1 1 22-DEC-1992 10:48
- [X1.X2.X3.X4.X5]X6.DIR;1 1 22-DEC-1992 10:48
- [X1.X2.X3.X4.X5.X6]X7.DIR;1 1 22-DEC-1992 10:48
- [X1.X2.X3.X4.X5.X6.X7]X8.DIR;1 1 22-DEC-1992 10:48
- [X1.X2.X3.X4.X5.X6.X7]X9.DIR;1 1 22-DEC-1992 10:50
- [X1.X2.X3.X4.X5.X6.X7.X9]X10.DIR;1 1 22-DEC-1992 10:50
- [000000]ZZ.DAT;1 1 22-DEC-1992 11:01
- []000000.DIR;1 2 9-DEC-1992 07:39
- []TEST1.DAT;1 1 22-DEC-1992 11:01
- []X11.DIR;1 1 22-DEC-1992 10:50
- []X12.DIR;1 1 22-DEC-1992 10:50
- []X13.DIR;1 1 22-DEC-1992 10:50
- []X14.DIR;1 1 22-DEC-1992 10:50
- []X15.DIR;1 1 22-DEC-1992 10:50
- []X16.DIR;1 1 22-DEC-1992 10:50
- []XXXX.XXXX;1 1 22-DEC-1992 10:50
-
- Total of 35 files, 1220 blocks
- End of save set
-
- **************************************
- NOTE: All files and directories will not be restored correctly if a disk
- problem occurs.
-
-
- >
- >You can also create files which, when browsed (say, with a $ TYPE filename
- >command), will result in the immediate termination of the browsing process!
- >(Make sure BUGCHECKFATAL is FALSE). Such a file cannot be COPYed, although
- >it can be BACKUPped. PRINTing such a file will cause the queue to stop.
- >{sudden thought, I wonder what would happen if this poison file were a valid
- >directory format, and I renamed it to .DIR.1? Another avoidance technique!}
-
- Cool... how do you do this?
-
-
- >
- >If you have a MAIL subdirectory, then having extra files in there with names
- >like MAILbigfilename.MAI is fairly inconspicuous. Unless somebody is being
- >particularly snoopy, (or using the $ SEARCH command), then this method is a
- >decent "avoidance" technique.
- >
-
- Now that's a creative way to do it. Just hope MAIL doesn't decide to use
- that MAILbigfilename.MAI sometime - though it may check for this and handle
- it correctly, I've no idea how to test this. (besides I'm sure that the
- possibility is quite small)
-
- Anyway, just wanted to make sure people are aware of problems with using
- some of the "avoidance" techniques.
-
-
- --
- -----
- Jeff Sue
- - All opinions are mine - (and you can't have any, nya nya nya)
-