Content is it worth defraging

[peter12345]

I have a Kindle 3 Keyboard - would it make it run more stable, and improve battery performance.

Also I have only 120MB free from 3 Giga, how much free space is best?

:}

[ixtab]

Originally Posted by peter12345

I have a Kindle 3 Keyboard - would it make it run more stable, and improve battery performance.

Also I have only 120MB free from 3 Giga, how much free space is best?

:}

IMO, defragmenting doesn't make a lot of sense -- though it probably won't hurt either. (see 3 posts down)

In terms of "how much free space is best": I don't understand the question The entire purpose of storage devices is to store data. You don't gain anything by keeping a disk empty. It won't magically be faster or anything -- you'll just have less meaningful data on it.

120 MB is still a lot of space if all that you want to store is a couple of books.

[peter12345]

This is a very fast response.

:} Though the free space is also needed to defrag. I am sorry I am a little mad about defrag - I would defrag my own brain if I could connect it to my laptop, though I think u would only find several folders there :}

[thomass]

HTG Explains: Why Linux Doesn’t Need Defragmenting

[ixtab]

Originally Posted by thomass

HTG Explains: Why Linux Doesn’t Need Defragmenting

Good link - but we're talking about the FAT 32 partition here

Then again, from the article:

[...]this is different for solid state drives, which don’t have moving parts and shouldn’t be defragmented – defragmenting an SSD will actually reduce its life.

I'm not a hardware expert, but I would assume that the same is true for the Kindle "drive".

[JeremyR]

For the longest time, people said you shouldn't defrag flash memory. But lately I've been reading that it does actually improve performance, and have done tests to show it does.

Flash memory does have a limited number of reads/writes, but I have to think the various other parts of the Kindle will wear out first.

With that said, I think it's just the number of files that slows down the Kindle, not so much the accessing them.

And while the Kindle uses Linux, doesn't it use the Fat 32 file system?

[ixtab]

Originally Posted by JeremyR

And while the Kindle uses Linux, doesn't it use the Fat 32 file system?

https://wiki.mobileread.com/wiki/Kind...ing#Partitions

(that's for the Touch/Paperwhite, but the general layout is the same on other models, just with different sizes)

[crich70]

Originally Posted by peter12345

This is a very fast response.

:} Though the free space is also needed to defrag. I am sorry I am a little mad about defrag - I would defrag my own brain if I could connect it to my laptop, though I think u would only find several folders there :}

The problem with the OISRS (Organic Information Storage & Retrieval System) is that it is very buggy. You start out with an almost empty unit hooked up to several other systems into which you must slowly input data and you can't even be sure the data will be retained during the 1st decade or two of operation. By the time you know how well the OISRS is going to work the warranty has almost certainly expired, and you find yourself locked out of the system, which by now has engaged a self-program mode which no one told you about when you 1st ordered the unit. It's a real mess.

[Mike L]

Quote: HTG Explains: Why Linux Doesn’t Need Defragmenting

Good link - but we're talking about the FAT 32 partition here

Surely this is not a question of Linux vs FAT 32, but rather flash drive vs traditional hard drive (the kind that spins).

The purpose of defragging is to move file clusters close to each other, so as to minimise movement of the read/write head. But flash drive don't have a read/write, and there is no benefit in making the files contiguous. It's just as fast to read two clusters that are "far apart" and ones that a contiguous.

At least, that's my understanding. I'm not a hardware expert.

Mike

[discopig]

http://www.wizcode.com/articles/comm...ths_and_facts/

http://ask-leo.com/should_i_defragme...ash_drive.html
seems there's contradicting evidence on both sides...

[JeremyR]

Well, just out of curiosity, I defragged my K3 (er, Kindle Keyboard).

It was 29% fragmented after I guess almost 3 years of use. Took surprisingly long to defrag, almost 2 hours.

I want to say it feels quicker after I defragged it, but it could be my imagination. Should have timed things before I did.

But here is a test someone did on an SD card

http://www.lagom.nl/misc/flash_fragmentation.html

Quite a noticeable difference - more than twice as fast to read

[hawhill]

I doubt you will notice a difference on the Kindle. The device is slow overall, but fast enough to do its job. That said: You might get a small speed gain, yes. SD/MMC blocks (the built-in flash is a eMMC, so it's more like an SD-Card than like a modern SSD) are quite small and the protocol has means for reading consecutive blocks without issuing new commands. But then the overhead for reading individual blocks isn't that big either. And it won't need a lot of CPU power at that point as the hardware has support for SD/MMC storage anyway. Having you flash erase/write cycles eaten up by a "defrag" doesn't sound like a good idea overall. Especially so as the MMC is masking the real flash layout, like the erase block size, for which the access could be optimized otherwise.

[peter12345]

Yes it can take 2 hours to defrag, but the second time you defrag it just takes a few seconds. It did feel that little bit more smoother afterwards....

thanks for the link:

http://www.lagom.nl/misc/flash_fragmentation.html

so if it is twice as fast battery life will increase too

also our brains are defraged when we go to sleep in dream mode - that is why we have so many subjective modes - my theory