Talk:SpinRite

From Wikipedia, the free encyclopedia

Contents 1 Disputed? 2 Versions 5 vs. 6 3 The way it works 4 like everything out of grc; it's odd 5 Thoughts 6 Linux tools 7 IDE Feature Disable during Dynastat Recovery

[edit] Disputed?

I've always thought (and been told by people who know better than I) that SpinRite was scamware - see the link to criticism at the bottom for more. This article, for the most part, reads like a press release without mentioning the criticisms except in passing. I'm adding the disputed tag although maybe npov would be more appropriate. Discuss! Brianski (talk) 06:20, 24 December 2007 (UTC)

Calling a helpful and well known work of software that is sold directly at a fair price a "scam" isn't much of a "neutral point of view" either. All tools may not work for every job. Spinrite is a professional recovery tool.

While it may be true that some people criticize SpinRite and refer to it as scamware, you'd need to identify exactly why it's scamware, and then either test spinrite to see if it was true, or find someone who already has and reference that material.

Otherwise you're putting up the disputed tag because "some guy doesn't like it". —Preceding unsigned comment added by 68.12.87.209 (talk) 18:48, 12 January 2008 (UTC)

"I've always thought (and been told by people who know better than I) that SpinRite was scamware - see the link to criticism at the bottom for more" - I have read the link twice. The link is a response to a posting on a newsgroup by Steve Gibson. Steve Gibson was the founder of Spinrite. Most of the link is the response of John Navas who wanted to give the facts (his own words) on the newsgroup. Read his claims of facts very carefully because he does talk about personal disagreements between them and gives his own personal opinions about the durability and manufacturing of hard drives. I don't see any test data on the Spinrite product or on hard drives. Since Brianski is stating his thoughts on Spinrite, and the oral communications of people who know better than him, and the newsgroup posting does not show any facts I submit that all of the arguments for the disputed tag are personal opninion. I don't even agree with including the reference link in the article. The disputed tag is not a correct tag. Perhaps there is a better tag (neutrality, or one involving comercialism)? I would be happy to continue this discussion as it relates to improving the article or whether the dispute tag should be included. If there is no further discussion I may choose to remove the disputed tag. I state here that I am not involved in the sale, manufacturing, advertising or in any other form with Gibson Research. I do not own this software. I did use this software on a PC about 15 years ago. The hard drive was not crashed, and I used it more as a preventive method. I probably used the product on roughly a weekly basis for 2-3 years. It did marks some hard drive clusters as "bad" and moved data to unused clusters. I kept no product testing records because I was only a consumer of the product, not testing it. I have no idea how it works (then or now). I state these things because I do not have anything personal gain if the tag is removed nor do I have any persoanl gain if the tag remains. I would like to see the justification for the tag as a Wikipedia editor. Mfields1 (talk) 08:34, 8 March 2008 (UTC)

[edit] Versions 5 vs. 6

Version 6 seems rather different from Version 5. Version 6 seems faster, more practical for modern larger hard drives. Version 5 was perhaps more thorough. Version 6 is supposed to work with "all" partition types, whereas Version 5 required working, common partition formatting, such as FAT. Version 6 seems to be more oriented to the raw hard drive, as opposed to the formatted partitions. Version 6 documentation may still not be completed.

It usually takes hours to scan a hard drive with SpinRite. In some cases it can take days.

A hard disk with bad blocks can be hard to scan with SpinRite 5, because of difficulty formatting to prepare for SpinRite. A ghost image of an empty formatted partition can be used to "fake format" the partition, so that SpinRite can scan it.

SpinRite 5 always insists on scanning each partition from the beginning, and can get very bogged down on bad areas. Is there any good quick tool for getting a picture of where all the bad areas are?

Both versions may interact in mysterious ways with the undocumented proprietary hidden hard drive features that manage bad block mapping. SpinRite may sometimes be able to recover blocks that had been mapped out? Some say this is a dangerous feature.

If you wish to re-use hard drives with significant bad areas, the best thing is to re-partition and stay completely away from those areas. However, this must be done manually with awkward (non-Microsoft) partitioning utility tools. There seem to be no tools that make it easy to scan for bad areas and then re-partition to avoid them. 69.87.193.26 13:34, 24 May 2006 (UTC)

GNU ddrescue may be a better way of not getting bogged down in bad sectors, as you can limit or turn off retries. You can simply use it to copy to /dev/null to get a picture of bad areas, which are recorded concisely in its log file.

SpinRite only recovers bad blocks to 'good', which I agree is very dangerous, at level 5. Using it at level 2, it will only rewrite data recovered from sectors that are bad but just-readable with DynaStat, and of course the hard drive will automatically re-map those to its good spare sectors. While this is a bit dangerous, I think it's a valid option for people who don't have a spare hard disk available and need to try to recover as much data as possible.

The ideal, if you have a spare hard disk of the right size is to first run GNU ddrescue, which will extract as much data as possible from the disk without any disk writes, and then run SpinRite at level 2 to attempt to recover more data with some rewrites. Then you can run GNU ddrescue again, and copy the additional recovered data - it automatically merges recovered data into the copy of the disk.

Richard Donkin 09:06, 19 November 2006 (UTC)

[edit] The way it works

The page states:

When it encounters a block which is hard to read, it repeatedly attempts to re-read it, and tries to determine the value of each byte. The data is then saved onto the same disk (after re-allocating the physical block) which is a potentially risky operation if the disk write head is not operating properly.

I used the program several times, and studied enough the documentation, it is not the way the program works, the data are not simply re-allocated, instead, a block is read, then it is written with an opposite sequence of data (if it was 00110101, it becomes 11001010), then it is rewritten again with the original sequence; it is done even several times (hopefully I am not violating any WP or copyright policy in disclosing such information).Dr. Who 00:41, 17 August 2006 (UTC)

[edit] like everything out of grc; it's odd

For a program that's worked wonders for many, for over a decade, it still surprises me that gibson has to resort to such bs pseudoscience to get it to sell. I'd of thought reputation alone would be enough... be he doesn't think so. Funny. -- BesigedB ^(talk) 22:55, 16 September 2006 (UTC)

[edit] Thoughts

Im running SpinRite right now on a Samsung disk with atleast 9 bad blocks. The DynaStat Data Recovery has some nice graphics, but I am very skeptical to basically all the claims made by it (the graphics and its website). I heard from one of my friends that it was "awesome" basically, and could really repair disks. I can tell you it did not work on my disk though.

Anyway, I am interested in how it gets the data from the disk: does ATA return data even if the checksums don't match up or does his program actually do anything special? I also wonder why the regular re-allocation algorithms dont work as good as SpinRite, if thats true. Ive tried running a bunch of the recovery tools from Ultimate Boot CD, but none were able to "repair" or reallocate the sectors. It seems SpinRite isnt either, though, so.. 62.84.192.238

I believe that ATA/IDE disks return data even if it fails ECC checks, at least when driven in the way SpinRite does. Richard Donkin 09:01, 19 November 2006 (UTC)

SpinRprite has worked for me. I had problems with a laptop 2.5 inch hdd. There were clicking noises coming from the disk. At first the drive would suddenly slow way down during read/write operations, and later on Windows would hang at boot. I ran spinrite, which claimed that the drive was dying, and it found 3 bad sectors, 2 of which it could not recover. After this windows went past the booting process, but claimed that some system files were missing. copied all the Data and got a new HDD. SpinRite, even though it has some false advertising, works in some cases...--169.232.119.69 (talk) 23:14, 11 January 2008 (UTC)

[edit] Linux tools

The discussion of Linux alternatives is very, very useful, though it should really be spun off into another page with a pointer from here. I recently had the misfortune of needing to use all these tools, including SpinRite and the Linux dd_rhelp and GNU ddrescue tools, and made an update to SpinRite accordingly, based on some information I found elsewhere. I think that the page should talk about the benefits of GNU ddrescue, which is much faster than dd_rhelp - if this had been mentioned I wouldn't have wasted time getting dd_rhelp and dd_rescue working. Richard Donkin 09:01, 19 November 2006 (UTC)

[edit] IDE Feature Disable during Dynastat Recovery

One of the features of Spinrite 6, which seems to have been missed, is that IDE sector sparing (and other features) are disabled during data recovery process. This prevents the drive from automatically replacing the bad sector with a good one from the spare sectors shipped on all drives when data read errors occur.

In the Dynastat process, which tries repeatedly to re-read the bad sector by approaching the sector from different positions and applying statistical analysis on the read data, sector sparing is disabled preventing the drive from giving up to early.

I don't believe that other similar tools do this, and this seems to be one of the reasons that Spinrite is successful at recovering data on drives reporting errors.

Spinrite is not always successful but when it works it works well.

The problem is spinrite doesn't say how it does this. I think that only certain drives provide an interface for this. But spinrite just claims it can do this whenever it likes to whatever drives. —Preceding unsigned comment added by 155.144.251.120 (talk) 22:33, 28 January 2008 (UTC)