I have a couple older distros of SuSE -- 6.4 & 7.2 and was wondering if there is an easy way to do it with a journaling FS? Whenever I'd get a hit/crash in ext2, it would sorta eat itself up, and then over a period of time, become unuseable. 6.4 has a wide variety of packages no longer available on the new distros, and I'd like to have a drive set up so I can boot it when wanted.
Gary HIldebrand St. Joseph, MO
While I'm sure there will be a lot of votes for ReiserFS, lets get real folks. Ext3 is being developed by a bunch of people who are free to leave where they live whenever they want, use a computer without supervision, and also not have their wages garnished by the state to pay for their living arrangements.
Hans Reiser, the developer of the Reiser filesystem, murdered his wife Nina Reiser about 2 years ago (to avoid hefty child support payments), was convicted of the crime, has already shown authorities where he buried her body, and is about to start a 15-years-to-life prison term for that murder. As the primary developer of the Reiser filesystem, and owner of the Namesys company which develops ReiserFS (which is unable to pay any employee's salary while he is incarcerated), Reiser4 development ground to a halt during his trial, and is likely to simply die off.
So the fact is that if your choice is between Reiser and Ext3, you may want to go with Ext3 simply because Ext3 will receive regular updates, and ReiserFS is dead in the ground (much like Nina Reiser).
--- On Tue, 9/23/08, gary hildebrand [email protected] wrote:
I have a couple older distros of SuSE -- 6.4 & 7.2 and was wondering if there is an easy way to do it with a journaling FS? Whenever I'd get a hit/crash in ext2, it would sorta eat itself up, and then over a period of time, become unuseable. 6.4 has a wide variety of packages no longer available on the new distros, and I'd like to have a drive set up so I can boot it when wanted.
Y'know, you could have just said the following: "You may want to stay away from ReiserFS because the lead developer is now incarcerated, and it may fall into disrepair." All of the other stuff is really immaterial to a technical decision. :)
I personally would add that ext3 is in much more common use and thus may be more supportable. It also has the added benefit that you can easily convert an existing ext2 filesystem to ext3 one, which may be desirable to the original poster.
Jeffrey.
On Tue, Sep 23, 2008 at 3:13 PM, Leo Mauler [email protected] wrote:
While I'm sure there will be a lot of votes for ReiserFS, lets get real folks. Ext3 is being developed by a bunch of people who are free to leave where they live whenever they want, use a computer without supervision, and also not have their wages garnished by the state to pay for their living arrangements.
Hans Reiser, the developer of the Reiser filesystem, murdered his wife Nina Reiser about 2 years ago (to avoid hefty child support payments), was convicted of the crime, has already shown authorities where he buried her body, and is about to start a 15-years-to-life prison term for that murder. As the primary developer of the Reiser filesystem, and owner of the Namesys company which develops ReiserFS (which is unable to pay any employee's salary while he is incarcerated), Reiser4 development ground to a halt during his trial, and is likely to simply die off.
So the fact is that if your choice is between Reiser and Ext3, you may want to go with Ext3 simply because Ext3 will receive regular updates, and ReiserFS is dead in the ground (much like Nina Reiser).
--- On Tue, 9/23/08, gary hildebrand [email protected] wrote:
I have a couple older distros of SuSE -- 6.4 & 7.2 and was wondering if there is an easy way to do it with a journaling FS? Whenever I'd get a hit/crash in ext2, it would sorta eat itself up, and then over a period of time, become unuseable. 6.4 has a wide variety of packages no longer available on the new distros, and I'd like to have a drive set up so I can boot it when wanted.
Anybody good with writing user-space filesystem drivers? I would like to build a file system on top of SQLite with the fts3 module. I think Unix semantics can be delivered with just a few tables.
The files table: BLOB data docid
a directories table, containing nothing but a rowid rowid
a directory entries table: integer directory text type text name integer entity
On Tue, Sep 23, 2008 at 3:32 PM, David Nicol [email protected] wrote:
Anybody good with writing user-space filesystem drivers? I would like to build a file system on top of SQLite with the fts3 module. I think Unix semantics can be delivered with just a few tables.
The files table: BLOB data docid
a directories table, containing nothing but a rowid rowid
a directory entries table: integer directory text type text name integer entity
Surely this is a good starting point:
"The libsqlfs library implements a POSIX style file system on top of an SQLite database. It allows applications to have access to a full read/write file system in a single file, complete with its own file hierarchy and name space. This is useful for applications which needs structured storage, such as embedding documents within documents, or management of configuration data or preferences. Libsqlfs can be used as an shared library, or it can be built as a FUSE (Linux File System in User Space) module to allow a libsqlfs database to be accessed via OS level file system interfaces by normal applications. "
http://www.nongnu.org/libsqlfs/
Justin Dugger
On Tue, Sep 23, 2008 at 3:41 PM, Justin Dugger [email protected] wrote:
Surely this is a good starting point:
Yes, thanks. I want to talk about extending that work to use the FTS3 indexing, making it a kernel module that uses a raw device rather than a file, and file system interface for making search queries.
--- On Tue, 9/23/08, Jeffrey Watts [email protected] wrote:
On Tue, Sep 23, 2008 at 3:13 PM, Leo Mauler [email protected] wrote:
While I'm sure there will be a lot of votes for ReiserFS, lets get real folks. Ext3 is being developed by a bunch of people who are free to leave where they live whenever they want, use a computer without supervision, and also not have their wages garnished by the state to pay for their living arrangements.
Hans Reiser, the developer of the Reiser filesystem, murdered his wife Nina Reiser about 2 years ago (to avoid hefty child support payments), was convicted of the crime, has already shown authorities where he buried her body, and is about to start a 15-years-to-life prison term for that murder. As the primary developer of the Reiser filesystem, and owner of the Namesys company which develops ReiserFS (which is unable to pay any employee's salary while he is incarcerated), Reiser4 development ground to a halt during his trial, and is likely to simply die off.
So the fact is that if your choice is between Reiser and Ext3, you may want to go with Ext3 simply because Ext3 will receive regular updates, and ReiserFS is dead in the ground (much like Nina Reiser).
Y'know, you could have just said the following: "You may want to stay away from ReiserFS because the lead developer is now incarcerated, and it may fall into disrepair." All of the other stuff is really immaterial to a technical decision. :)
Kevin Mitnick was also once incarcerated, even to the point of being denied use of any electronic equipment more complex than a landline phone while on parole, but now runs his own computer security company. Given the prior example of Kevin Mitnick, I felt that the general word "incarceration" was inadequate to explain just how incredibly dead Reiser4 has become.
I personally would add that ext3 is in much more common use and thus may be more supportable. It also has the added benefit that you can easily convert an existing ext2 filesystem to ext3 one, which may be desirable to the original poster.
I think you missed my point, but whatever. ReiserFS is open source - just because the guy is in prison doesn't mean the filesystem is "incredibly dead". I certainly agree that losing its chief advocate (and original author) poses a serious problem for the development effort, but many other open source projects have had similar losses of leadership in the past and have yet survived. My point was that the "Nancy Grace" extras really don't have much to do with the software itself.
Regardless, I personally recommend that folks use ext3 unless they have special, high-end requirements. I'm not sure if the original poster's version of SuSE supports ext3 (I've never followed SuSE closely), but if it doesn't I'd recommend that he upgrade to a version that does.
Jeffrey.
On Fri, Sep 26, 2008 at 12:05 AM, Leo Mauler [email protected] wrote:
Kevin Mitnick was also once incarcerated, even to the point of being denied use of any electronic equipment more complex than a landline phone while on parole, but now runs his own computer security company. Given the prior example of Kevin Mitnick, I felt that the general word "incarceration" was inadequate to explain just how incredibly dead Reiser4 has become.
--- On Fri, 9/26/08, Jeffrey Watts [email protected] wrote:
ReiserFS is open source - just because the guy is in prison doesn't mean the filesystem is "incredibly dead". I certainly agree that losing its chief advocate (and original author) poses a serious problem for the development effort, but many other open source projects have had similar losses of leadership in the past and have yet survived.
The mere existence of open source code is like DNA in a jar in a freezer: a blueprint for someone to take and run with, but not actual evidence of life, and until someone gets out the metaphorical "turkey baster" and finds a not-so-metaphorical willing developer, the project will remain DNA in a jar.
Case in point: *Gentoo Linux*, which is dying a fairly swift death thanks to the departure of Gentoo Linux's project leader.
Gentoo's decline: A case of missing leadership September 22, 2008 7:07 AM PDT
http://news.cnet.com/8301-13505_3-10047439-16.html?tag=mncol;title
TinyURL: http://tinyurl.com/49b3th
"The most interesting thing about the current state of gentoo is that it's a very clear (and well documented) example of how the success of a large open source project, regardless of the personal devotion of its user base, is tightly coupled to the strength of its leadership. Interesting also that despite the projects strong attraction of 'power users', the community has been unable to convert these users into active developers."
I think this pretty much underlines my point that, absent proper leadership, an open source project is a bunch of bits on a hard drive.
Blog Entry from the creator of Gentoo, who saw Gentoo's demise coming last year: http://blog.funtoo.org/2007/07/so-can-i-have-gentoo-back.html
"It appears that the Gentoo Foundation may be disappearing in a matter of time. Apparently, no one has time to actually do the work required to run the Foundation."
Hans Reiser had to pay a half dozen people to do work on his ReiserFS. With Reiser in prison no one is getting paid for their work, so there's no real incentive to improve on it.
Leo, I totally get all of that. My comment was that the gossipy extras wasn't really germane, and served as a distraction to a _technical discussion_.
Here's the problem with your attempt to compare Gentoo with ReiserFS. Gentoo is a _distribution_. If the Gentoo development team doesn't update it regularly, it usually becomes rapidly vulnerable to security issues and becomes unsafe. That is the main problem with distributions becoming abandonware.
ReiserFS is a filesystem. SuSE supports ReiserFS. Just because Hans Reiser is in jail doesn't mean his filesystem will start losing data or murdering russians. Last I looked, it should continue to work just as before. SuSE, as the _distribution_, will continue to maintain ReiserFS as best it can. I can imagine that if SuSE starts having issues with the software that they can't fix themselves they might drop SuSE support in the future (but not in the present). I also imagine that ReiserFS support will be dropped should the project dry up and no new development occur (die on the vine). But none of that has any effect on his ability to safely run ReiserFS, today.
However, my point to you is that the sordid details have very little to do with a _technical decision_ as to which filesystem to use. It doesn't sound like he's running a datacenter. He's looking for advice on a journalling filesystem. I think it's germane to offer criticism of the downsides of ReiserFS, and I think it can be informative to point out that there's a future development concern, but he's asking about a journaling filesystem on an older distribution release, so I don't think he's too worried about the future.
Regardless of all of this, consensus is probably that he ought to use ext3, and that if he's not on a distro that supports it (I don't know my SuSE releases anymore), he ought to upgrade to one that does.
Jeffrey.
P.S. I'm really not interested in getting into another endless debate with you Leo, so you may want to save some of your time by keeping your inevitable response short. I'm unlikely to reply.
On Fri, Sep 26, 2008 at 2:30 AM, Leo Mauler [email protected] wrote:
The mere existence of open source code is like DNA in a jar in a freezer: a blueprint for someone to take and run with, but not actual evidence of life, and until someone gets out the metaphorical "turkey baster" and finds a not-so-metaphorical willing developer, the project will remain DNA in a jar.
Case in point: *Gentoo Linux*, which is dying a fairly swift death thanks to the departure of Gentoo Linux's project leader.
Gentoo's decline: A case of missing leadership September 22, 2008 7:07 AM PDT
http://news.cnet.com/8301-13505_3-10047439-16.html?tag=mncol;title
TinyURL: http://tinyurl.com/49b3th
"The most interesting thing about the current state of gentoo is that it's a very clear (and well documented) example of how the success of a large open source project, regardless of the personal devotion of its user base, is tightly coupled to the strength of its leadership. Interesting also that despite the projects strong attraction of 'power users', the community has been unable to convert these users into active developers."
I think this pretty much underlines my point that, absent proper leadership, an open source project is a bunch of bits on a hard drive.
Blog Entry from the creator of Gentoo, who saw Gentoo's demise coming last year: http://blog.funtoo.org/2007/07/so-can-i-have-gentoo-back.html
"It appears that the Gentoo Foundation may be disappearing in a matter of time. Apparently, no one has time to actually do the work required to run the Foundation."
Hans Reiser had to pay a half dozen people to do work on his ReiserFS. With Reiser in prison no one is getting paid for their work, so there's no real incentive to improve on it.
--- On Fri, 9/26/08, Jeffrey Watts [email protected] wrote:
ReiserFS is a filesystem. SuSE supports ReiserFS.
As a secondary option only, and they were paying Hans Reiser's now nearly-defunct company Namesys to support ReiserFS for them (at Hans Reiser's insistence).
Just because Hans Reiser is in jail doesn't mean his filesystem will start losing data or murdering russians. Last I looked, it should continue to work just as before.
Reiser3 was already having a lot of problems, and Reiser4 was still vaporware:
http://lists.opensuse.org/opensuse-factory/2006-09/msg00542.html
SuSE, as the _distribution_, will continue to maintain ReiserFS as best it can.
Actually they haven't been supporting ReiserFS. They've been paying Reiser's company, Namesys, for ReiserFS support, but with Hans Reiser in prison (thanks to all the "gossipy details"), thats ended.
Of course, the main point should be that SuSE has been backing away from ReiserFS since before the "gossipy details", switching to Ext3 as its default filesystem about two years ago:
October 12, 2006 5:30 PM PDT Novell makes file storage software shift By Stephen Shankland Staff Writer, CNET News
"Novell is changing the file system software used by default in its Suse Linux operating system, aligning with rival Red Hat and moving away from a project whose future has become entangled with the fate of a murder suspect."
http://news.cnet.com/Novell-makes-file-storage-software-shift/2100-1016_3-61...
TinyURL: http://tinyurl.com/53pomq
However, my point to you is that the sordid details have very little to do with a _technical decision_ as to which filesystem to use.
I will continue to think that limiting the descriptive language to the vague phrase "incarceration of the lead developer" is about as inadequate as saying that the current tendency to "lie down on his back a lot" will prevent comic George Carlin from doing any more shows. The details *matter*.
"The most interesting thing about the current state of gentoo is that it's a very clear (and well documented) example of how the success of a large open source project, regardless of the personal devotion of its user base, is tightly coupled to the strength of its leadership.
On Fri, Sep 26, 2008 at 6:50 PM, Leo Mauler [email protected] wrote:
Actually they haven't been supporting ReiserFS. They've been paying Reiser's company, Namesys, for ReiserFS support, but with Hans Reiser in prison (thanks to all the "gossipy details"), thats ended.
I don't think you're making a fundamental mistake here. Not all distributions are the same.
Gentoo is a ricer's OS. Folks who use it like to tweak systems and learn stuff, and maybe put some bitchin' neon ground effects on it and a killer stereo. SuSE is a professional company that sells guaranteed products.
If some part of Gentoo doesn't work well, well then tough cookies there Johnny Rocket. Your mileage may vary. No warranty. No refunds.
SuSE guarantees their product. If a supported version of SuSE supports ReiserFS, then the company will support it to the best of their ability.
I don't care about the rest of the crap you're talking about, as it's not what the original poster asked about. He wanted to know of a journaling filesystem for his old SuSE boxes. Someone suggested ReiserFS. I, like most others, recommended he upgrade and use ext3. However should he choose to use his older OS and should ReiserFS be available and supported, it may not be a bad idea _technically_.
If Richard Stallman went out and murdered some kittens would you stop using the GCC compiler? Would you shun FSF projects? I think that somehow GCC will still be a _technically_ sound compiler regardless of it's original author's penchant for cat-butchery. Perhaps its long term future might be in jeopardy, but I'd think it'd still be able to compile my hello.c on my current OS.
You are okay in pointing out that ReiserFS might not be a good long term strategy but what does any of that have to do with him using ReiserFS on an old distribution where it is _fully supported by a professional company_? This isn't the Department of Homeland Security here. I don't think this guy's choice of filesystems has to be that big a deal.
I suppose the worst part is that I know I'm talking to a tree here. You're going to go over your same arguments again and again, and you'll probably Google even more factoids to back up your obtuse position. I've already responded twice too many times, and in the interest of sparing others another NetNews-like discussion between us I'll let you respond and then I'll give it a rest.
Ciao. Jeffrey.
--- On Sat, 9/27/08, Jeffrey Watts [email protected] wrote:
On Fri, Sep 26, 2008 at 6:50 PM, Leo Mauler [email protected] wrote:
Actually they haven't been supporting ReiserFS. They've been paying Reiser's company, Namesys, for ReiserFS support, but with Hans Reiser in prison (thanks to all the "gossipy details"), thats ended.
I don't think you're making a fundamental mistake here. Not all distributions are the same.
Gentoo is a ricer's OS. Folks who use it like to tweak systems and learn stuff, and maybe put some bitchin' neon ground effects on it and a killer stereo. SuSE is a professional company that sells guaranteed products.
If some part of Gentoo doesn't work well, well then tough cookies there Johnny Rocket. Your mileage may vary. No warranty. No refunds.
SuSE guarantees their product. If a supported version of SuSE supports ReiserFS, then the company will support it to the best of their ability.
Which won't be much now that Namesys' support has stalled. SuSE has been encouraging users to use Ext3 over ReiserFS for some time now, for quite a number of very good reasons, starting well before the Hans Reiser incarceration.
Someone probably came into the SuSE boardroom and said something like "REISERFS IS DYING BECAUSE IT'S NO LONGER A GOOD TECHNOLOGY. YOU OUGHT TO CONSIDER MOVING TO A NEWER TECHNOLOGY. I HEAR THERE IS THIS THING OUT THERE CALLED 'EXT3'."
After all, ReiserFS has a lot of technical problems, and Reiser4 will likely never see the light of day. Technology which is no longer good must be dropped, Jeffrey Watts, as you told me over and over a few months ago. Nevermind if some people still use it, clearly they are idiots who must be re-educated into using a filesystem which, for example, isn't still dependent on the BKL and thus unable to scale well due to only using one core at a time of a multi-core processor. Or won't get confused during a fsck and make a corrupted filesystem into an even more corrupted filesystem. Or won't see a ReiserFS disk image on a ReiserFS filesystem and get confused as to which disk contents (filesystem or image) are the right ones to recover.
Reformatting a ReiserFS filesystem can sometimes result in data files from the previous contents reappearing on the now allegedly-wiped hard drive, a detail which folks interested in tax breaks for donating old machines, while simultaneously protecting their corporate secrets, would be dismayed to learn.
Meanwhile, Ext3 now has all of ReiserFS's features, very few of its problems, and a few features not present in ReiserFS (such as XIP).
I've deleted the rest of your crap (where you try to bait me into talking more about Hans Reiser's difficulties), which was irrelevant to the point that SuSE had already decided to stop using ReiserFS as its default filesystem before the Hans Reiser incarceration, due to ReiserFS' technical problems and the cost of getting all its ReiserFS support from Namesys. If you want to go on discussing a dead issue, be my guest, you seem to have developed an enjoyment of that sort of thing.
On Sun, Sep 28, 2008 at 11:58 AM, Leo Mauler [email protected] wrote:
--
Reformatting a ReiserFS filesystem can sometimes result in data files from the previous contents reappearing on the now allegedly-wiped hard drive, a detail which folks interested in tax breaks for donating old machines, while simultaneously protecting their corporate secrets, would be dismayed to learn.
Anyone who thinks that "reformatting a filesystem" is the same thing as "wiping a hard drive" doesn't understand the meanings of the words "reformatting", "wiping", or "filesystem' for that matter.
You can reformat an ext*, FAT or NTFS filesystem, but if you think you've "wiped" the filesystem, much less the drive, you are wrong beyond description. I can understand a non-technical user not understanding the distinction between "reformatting' and "wiping", but anyone working for a corporate IT department who can't tell them apart should be fired for incompetence.
If I'm asked to "wipe" someone's drive, I'll boot to a live CD or USB drive, get to a root prompt, and do something like
export pass=1 while [ $pass -lt 6 ] do printf '\nDrive Wipe Pass #%d' $pass dd if=/dev/random of=/dev/hda printf ' . ' dd if=/dev/zero of=/dev/hda printf ' . ' pass=`expr $pass + 1` done; printf '\nDONE. You may now turn off the computer'
This is not a theoretical exercise for me. I had a customer who was retiring, and wanted to be sure all data was removed from the server before it was delivered to the buyer. Since it was SCO OpenServer, instead of /dev/random, I used an outer loop of 3 passes with an inner loop that used /dev/byte/55, /dev/byte/ff, /dev/byte/aa, and /dev/byte/00 (after first creating the device nodes for 55 and aa), so that the entire HD was written to a dozen times, with the different patterns. At that point, I figured that there were maybe some guys at Langley (and their Russian and Chinese counterparts) who would have even a chance of recovering data from that drive.
I never referred to that operation as "(re)formatting".
Hi,
On Sun, Sep 28, 2008 at 08:10:37PM -0500, Monty J. Harder wrote:
Anyone who thinks that "reformatting a filesystem" is the same thing as "wiping a hard drive" doesn't understand the meanings of the words "reformatting", "wiping", or "filesystem' for that matter.
The D.O.D. doesn't use the term "wiping", they talk about "de-classifying" which requires "degaussing", not something easily done with software. I can't remember their term for "wiping", removing non-classified data from the media before selling to J. Q. Public, but the procedure was basically what you use for "wiping". They didn't seem to care much about the 0xAA and 0x55 patterns, they just wanted the 0's, 1's and random, and wanted thousands of passes rather than dozens of passes. And there was concern that the software might be doing all its writing to a disk buffer somewhere in RAM rather than the actual disk. So I guess if you want to "wipe" someone's disk you should first remove all but one stick of RAM, then fire up Open Office, the Gimp, and a few instances of Firefox, and THEN run your "wipe" program.
On Mon, Sep 29, 2008 at 2:24 AM, [email protected] wrote:
thousands of passes rather than dozens of passes. And there was concern that the software might be doing all its writing to a disk buffer somewhere in RAM rather than the actual disk. So I guess if you want to "wipe"
I don't believe that dd writes to a disk buffer in RAM. Certainly, the drive itself may use a write buffer, but writing the entire disk would easily exhaust that buffer. I don't know why they would want thousands of passes though. They may have had a point back in the days of MFM hard drives, but encoding schemes have become very sophisticated. Modern hard drives must employ extensive error correction codes just to be able to read the last written contents of a sector reliably, I stand by my contention that alternating 0s and random junk for a total of a dozen passes should render the data completely inaccessible to anyone (with the possible exception of those spy agencies, and I'm not even sure about them).
Well, he was talking about the government. They're always ten or more years behind, so I'm willing to bet that their disk wiping procedures probably DO date back from when there were MFM drives. ;-)
J.
On Mon, Sep 29, 2008 at 10:19 AM, Monty J. Harder [email protected]wrote:
I don't believe that dd writes to a disk buffer in RAM. Certainly, the drive itself may use a write buffer, but writing the entire disk would easily exhaust that buffer. I don't know why they would want thousands of passes though. They may have had a point back in the days of MFM hard drives, but encoding schemes have become very sophisticated. Modern hard drives must employ extensive error correction codes just to be able to read the last written contents of a sector reliably, I stand by my contention that alternating 0s and random junk for a total of a dozen passes should render the data completely inaccessible to anyone (with the possible exception of those spy agencies, and I'm not even sure about them).
--- On Mon, 9/29/08, Jeffrey Watts [email protected] wrote:
Well, he was talking about the government. They're always ten or more years behind, so I'm willing to bet that their disk wiping procedures probably DO date back from when there were MFM drives. ;-)
Seems to me the government is up to speed on this one.
"There is an abundant array of disk wiping applications available for computer users. The government standard is a medium security level application that specifies overwriting a hard drive six times through three iterations. Each iteration consists of two write-passes on a hard drive. The first iteration removes the files over at the drive surface, while the second iteration registers "zeros" on the surface. Finally, the government-designated code of 246 is assigned to a drive when the third iteration culminates. The only disadvantage of opting to wipe disk is the time it requires to do so."
http://ezinearticles.com/?Wipe-Disk-To-Rid-Off-Hard-Drive-Data&id=114307... TinyURL: http://tinyurl.com/4fy35d
When it comes to keeping things secret, the government seems to be up to speed. Its just when it comes to less important stuff like catching murderers (i.e., revealing secrets) when it falls behind.
On Mon, Sep 29, 2008 at 9:32 PM, Leo Mauler [email protected] wrote:
The government standard is a medium security level application that specifies overwriting a hard drive six times through three iterations. Each iteration consists of two write-passes on a hard drive. The first iteration removes the files over at the drive surface, while the second iteration registers "zeros" on the surface.
What the hell does that even mean?
"removes the files over at the drive surface"
It sounds like they say to write six times to the drive, with the even-numbered writes being 0s. I suggested twice that. So what do they recommend writing on the odd-numbered passes if not (pseudo)random junk?
Back when there was some correspondence between the data sent to the drive and the actual patterns written to disk, one could try to design a sequence of patterns to do a really good job of eliminating the "shadows" of previous writes. Since every drive potentially uses a different algorithm for the low-level storage, the logical thing to do would be to let the drive itself handle wiping.
Extend the command set to provide a directive to securely wipe a range of sectors on a drive. The drive would then implement its own method that takes into account the algorithm it uses. Since the drive has access to samples that are not passed through to the CPU, it would be able to tailor what it writes to what is on that sector, and after a few passes of read/write feedback, get things pretty thoroughly scrambled. In this instance, the drive would be writing patterns that it never writes to encode data, because it would be deliberately putting flux transitions between the normal locations where they would be located.
Also, when a drive detects that a sector is no longer reliable (even with the error-correction codes it can't read back what it just wrote to that sector), and is taken out of service (substituting a spare sector transparent to the CPU's knowledge) the retired sector should automatically receive this treatment, lest it contain sensitive info that could later be recovered by someone who bypasses the normal redirection.
Drives with this technology could be marketed as having "Secure Deletion" capabilities, and easily command premium prices. Wouldn't you gladly pay $10 more for a drive that can wipe sectors so well that even the spooks would get nothing out of them?
I can't figure out why the government just doesn't remove the hard drives and send them to a shredder. Even a private individual can achieve roughly the same level of security with a drill and a 1/2" drill bit. With the price of drives being what they are now it can't impact very much the resale value of the computer to sell it without a drive installed.
On Tue, Sep 30, 2008 at 11:29 AM, Monty J. Harder [email protected]wrote:
On Mon, Sep 29, 2008 at 9:32 PM, Leo Mauler [email protected] wrote:
The government standard is a medium security level application that specifies overwriting a hard drive six times through three iterations. Each iteration consists of two write-passes on a hard drive. The first iteration removes the files over at the drive surface, while the second iteration registers "zeros" on the surface.
What the hell does that even mean?
"removes the files over at the drive surface"
It sounds like they say to write six times to the drive, with the even-numbered writes being 0s. I suggested twice that. So what do they recommend writing on the odd-numbered passes if not (pseudo)random junk?
My understanding is that for classified stuff they in fact do that. The drives are "wiped" then sent to a contractor that melts them down.
Jeffrey.
On Tue, Sep 30, 2008 at 11:48 AM, Dana Smith [email protected] wrote:
I can't figure out why the government just doesn't remove the hard drives and send them to a shredder. Even a private individual can achieve roughly the same level of security with a drill and a 1/2" drill bit. With the price of drives being what they are now it can't impact very much the resale value of the computer to sell it without a drive installed.
On Tue, Sep 30, 2008 at 11:29 AM, Monty J. Harder [email protected] wrote:
Drives with this technology could be marketed as having "Secure Deletion" capabilities, and easily command premium prices. Wouldn't you gladly pay $10 more for a drive that can wipe sectors so well that even the spooks would get nothing out of them?
Absolutely not. I would be highly suspicious of using any such drive for my hypothetical criminal activity, as a drive marketed as having SD features might /really/ have some kind of covert channel phone-home-to-Echelon feature allowing the spooks to 0wn the box remotely. Perhaps interfacing with the pc-speaker-as-microphone feature.
There's a flaw with your reasoning.
You're assuming that Big Brother would put a backdoor of some kind into devices like that marked "Secure Deletion". However, if you're going to be Tin-Foil Hat Man you should be assuming that EVERY device has a Big Brother backdoor.
Thus your only real choice is to live in a cave and bang rocks together.
Jeffrey.
On Tue, Sep 30, 2008 at 1:52 PM, David Nicol [email protected] wrote:
Absolutely not. I would be highly suspicious of using any such drive for my hypothetical criminal activity, as a drive marketed as having SD features might /really/ have some kind of covert channel phone-home-to-Echelon feature allowing the spooks to 0wn the box remotely. Perhaps interfacing with the pc-speaker-as-microphone feature.
On Tue, Sep 30, 2008 at 01:58:56PM -0500, Jeffrey Watts wrote:
There's a flaw with your reasoning.
You're assuming that Big Brother would put a backdoor of some kind into devices like that marked "Secure Deletion". However, if you're going to be Tin-Foil Hat Man you should be assuming that EVERY device has a Big Brother backdoor.
Thus your only real choice is to live in a cave and bang rocks together.
Jeffrey.
Even there you have a problem. At least every time I have moved rocks, I uncovered a number of bugs. Even the rocks are bugged.
Thanks, -- Hal
you should be assuming that EVERY device has a Big Brother backdoor
What is wrong with that assumption (on mature technologies).
Thus your only real choice is to live in a cave and bang rocks together.
I see potential issues with that conclusion, unless you are really good at banging rocks (such that people will pay you).
Thanks,
Ron Geoffrion 913.488.7664
--- On Tue, 9/30/08, Jeffrey Watts [email protected] wrote:
On Tue, Sep 30, 2008 at 1:52 PM, David Nicol [email protected] wrote:
Absolutely not. I would be highly suspicious of using any such drive for my hypothetical criminal activity, as a drive marketed as having SD features might /really/ have some kind of covert channel phone-home-to-Echelon feature allowing the spooks to 0wn the box remotely. Perhaps interfacing with the pc-speaker-as-microphone feature.
There's a flaw with your reasoning.
You're assuming that Big Brother would put a backdoor of some kind into devices like that marked "Secure Deletion". However, if you're going to be Tin-Foil Hat Man you should be assuming that EVERY device has a Big Brother backdoor.
Thus your only real choice is to live in a cave and bang rocks together.
There's a flaw in your reasoning.
You're assuming that there's *no way* to prevent all of your computer peripherals from wirelessly transmitting data from the devices to a central government server.
However, that kind of assumption takes the mentality of someone who really does live in a cave and bang rocks together. Surely someone who claims to have computer knowledge has encountered the "metal sheets in the wall prevent home use of 802.11x wireless networking" problem before, right?
There's a very simple way of preventing wireless transmission: a Faraday cage, or just a Big Metal Room. Any criminal organization which doesn't get shut down repeatedly for stupidity has its most sensitive computers in just such a room, and all sensitive data is sneakernetted in and out of the room. A wired connection is even easier: unplug the sensitive computer's phone line and/or network connection.
Just build a home using sheet metal under the drywall, a sort-of "tinfoil house-hat".
Jeez Leo, you wouldn't know a joke if it hit you in the face. I'll make sure to use more smileys in the future.
J.
On Tue, Sep 30, 2008 at 5:27 PM, Leo Mauler [email protected] wrote:
There's a flaw in your reasoning.
You're assuming that there's *no way* to prevent all of your computer peripherals from wirelessly transmitting data from the devices to a central government server.
However, that kind of assumption takes the mentality of someone who really does live in a cave and bang rocks together. Surely someone who claims to have computer knowledge has encountered the "metal sheets in the wall prevent home use of 802.11x wireless networking" problem before, right?
On Tue, Sep 30, 2008 at 1:52 PM, David Nicol [email protected] wrote:
Absolutely not. I would be highly suspicious of using any such drive for my hypothetical criminal activity, as a drive marketed as having SD features might /really/ have some kind of covert channel phone-home-to-Echelon feature allowing the spooks to 0wn the box remotely. Perhaps interfacing with the pc-speaker-as-microphone feature.
Well, I wasn't making recommendations for criminal activity, but for protection against inadvertently disclosing sensitive material. In the real-world example I cited, it was a doctor that retired. Although HIPAA had not fully come into effect yet, he was rightly concerned about protecting confidential information entrusted to him by his patients. I can confidently say we fufilled his ethical and legal obligations.
If you're concerned about keeping your data safe from any government, particularly those of the United States, China, Russia, Israel, or an EU member, you need to take additional precautions, such as the use of full-drive encryption (TrueCrypt) and an open-source operating system that has enough eyeballs on it that there is little chance of a backdoor being slipped in somewhere. And then when you're done with the drive, electronically wipe it as I've described, drill a hole in it, smash it with a sledge hammer, and cast the parts into an active volcano like Mount Doom.
On Tuesday 30 September 2008 11:29:07 Monty J. Harder wrote:
Wouldn't you gladly pay $10 more for a drive that can wipe sectors so well that even the spooks would get nothing out of them?
No. If I'm ever worried about it, I'll use a sledge hammer.
No. If I'm ever worried about it, I'll use a sledge hammer.
Or you could try to mail the disk in question to yourself from Bogota and let Columbian customs destroy it for you (refers to yesterday's news item concerning Kevin Mitnick)
you could just `sync` to make sure it goes to the disk, but if I recall `dd` only writes to the device you tell it to and not some RAM location
On Mon, Sep 29, 2008 at 2:24 AM, [email protected] wrote:
Hi,
On Sun, Sep 28, 2008 at 08:10:37PM -0500, Monty J. Harder wrote:
Anyone who thinks that "reformatting a filesystem" is the same thing as "wiping a hard drive" doesn't understand the meanings of the words "reformatting", "wiping", or "filesystem' for that matter.
The D.O.D. doesn't use the term "wiping", they talk about "de-classifying" which requires "degaussing", not something easily done with software. I can't remember their term for "wiping", removing non-classified data from the media before selling to J. Q. Public, but the procedure was basically what you use for "wiping". They didn't seem to care much about the 0xAA and 0x55 patterns, they just wanted the 0's, 1's and random, and wanted thousands of passes rather than dozens of passes. And there was concern that the software might be doing all its writing to a disk buffer somewhere in RAM rather than the actual disk. So I guess if you want to "wipe" someone's disk you should first remove all but one stick of RAM, then fire up Open Office, the Gimp, and a few instances of Firefox, and THEN run your "wipe" program.
-- Jim _______________________________________________ Kclug mailing list [email protected] http://kclug.org/mailman/listinfo/kclug
--- On Sun, 9/28/08, Monty J. Harder [email protected] wrote:
On Sun, Sep 28, 2008 at 11:58 AM, Leo Mauler [email protected] wrote:
Reformatting a ReiserFS filesystem can sometimes result in data files from the previous contents reappearing on the now allegedly-wiped hard drive, a detail which folks interested in tax breaks for donating old machines, while simultaneously protecting their corporate secrets, would be dismayed to learn.
Anyone who thinks that "reformatting a filesystem" is the same thing as "wiping a hard drive" doesn't understand the meanings of the words "reformatting", "wiping", or "filesystem' for that matter.
Sorry, bad choice of words there, I do know what wiping means.
Actually, anyone who thinks that completely wiping a drive can be done with software is sorely mistaken. A friend of mine who, shall we say, did some "shady computer things" back in the 1980s, had an external hard drive (long before home computers had external hard drives, this was a homemade model) which was conveniently placed between two powerful electromagnets. Should anyone choose to come into his home to, shall we say, "inquire about his shady computer things", he could simply press a Big Red Button (his happened to be big and red) and completely wipe his hard drive.
There are government labs with the technology to go through a "software-wiped" drive and piece together some of what was previously on said "wiped" drive, but not drives which were "hardware-wiped". I wouldn't doubt that there are independent labs (such as in corporate secrets espionage) with the same level of technology.
I suppose ideally companies should buy brand new hard drives for their donated computers and run the old ones over with a steamroller.
Actually there's a new threat looming. The USB "thumbdrives" that we take for granted first gave the warning. Seems the "wear spreading" logics they use make for a grim risk that anything stored on one as being potentially recoverable... Which then made someone wake up sweating at midnight over a fruit phone's SSD. Oh? SSD? As in EEE etc. I suspect the confluence of EEE boxes being coveted by tech savvy folks and having been sold with XP may get interesting. As We all know that XP makes oddly redundant files. So someone getting a resold EEE and running Photorecovery or similar?
Back to Gentoo. To my recall in the past hardware not hefty enough took simply eternities to emerge on .The part I still am curious about centers on if the "creation" of a Gentoo image for a machine could now be done on a fast box then simply cloned to the lesser machine by dd.
On Mon, Sep 29, 2008 at 8:46 PM, Leo Mauler [email protected] wrote:
--- On Sun, 9/28/08, Monty J. Harder [email protected] wrote:
On Sun, Sep 28, 2008 at 11:58 AM, Leo Mauler [email protected] wrote:
Reformatting a ReiserFS filesystem can sometimes result in data files from the previous contents reappearing on the now allegedly-wiped hard drive, a detail which folks interested in tax breaks for donating old machines, while simultaneously protecting their corporate secrets, would be dismayed to learn.
Anyone who thinks that "reformatting a filesystem" is the same thing as "wiping a hard drive" doesn't understand the meanings of the words "reformatting", "wiping", or "filesystem' for that matter.
Sorry, bad choice of words there, I do know what wiping means.
Actually, anyone who thinks that completely wiping a drive can be done with software is sorely mistaken. A friend of mine who, shall we say, did some "shady computer things" back in the 1980s, had an external hard drive (long before home computers had external hard drives, this was a homemade model) which was conveniently placed between two powerful electromagnets. Should anyone choose to come into his home to, shall we say, "inquire about his shady computer things", he could simply press a Big Red Button (his happened to be big and red) and completely wipe his hard drive.
There are government labs with the technology to go through a "software-wiped" drive and piece together some of what was previously on said "wiped" drive, but not drives which were "hardware-wiped". I wouldn't doubt that there are independent labs (such as in corporate secrets espionage) with the same level of technology.
I suppose ideally companies should buy brand new hard drives for their donated computers and run the old ones over with a steamroller.
Kclug mailing list [email protected] http://kclug.org/mailman/listinfo/kclug
On Tuesday 23 September 2008, Leo Mauler wrote:
Hans Reiser, the developer of the Reiser filesystem, murdered his wife Nina Reiser about 2 years ago (to avoid hefty child support payments), was convicted of the crime, has already shown authorities where he buried her body, and is about to start a 15-years-to-life prison term for that murder. As the primary developer of the Reiser filesystem, and owner of the Namesys company which develops ReiserFS (which is unable to pay any employee's salary while he is incarcerated), Reiser4 development ground to a halt during his trial, and is likely to simply die off.
I'm not trying to justify murder, nor do I care or use ReiserFS in any revision, but with the abuse from both Nina and the state, I don't think Hans carries the full guilt of his crime nor should be portrayed in that light.
-
Dana Smith -
David Nicol -
gary hildebrand -
Geoffrion, Ron P [IT] -
Hal Duston -
Jeffrey Watts -
jim＠jimani.com -
Justin Dugger -
Leo Mauler -
Luke -Jr -
Monty J. Harder -
Oren Beck -
Philip Dorr