defraggler zero fill

[julevine]

i dont understand why  defragger  zero fill writes  0's to  drive  

should't  it be writing  binary  1's to pages and blocks

because how ssd programs  pages and blocks

Programmed block = binary value  0 

erased block = binary value 1 

or  is it  just to fill  pages and blocks with   data so after   it deletes the data  the ssd triggers trim  command on all these blocks 

but it it mostly for ssd's that dont support trim   

or  am i getting confused because  file disk view is  showing  0's   even though it a binary 1 

 

[Augeas]

I don't use Defraggler so I don't know what options it has but don't you mean CCleaner?

Whatever s/w it is the process of writing a file and then deleting it is totally different from erasing a NAND flash block. The SSD's grabage collection routines will erase the (invalid) block irrespective of what has been written there previously.

And don't bother to zero or anthing fill an SSD. Windows will not allow you to access TRIMed data, and the block erasure will remove anything that was written before permanently, so a normal delete will do.

[julevine]

talking about  defraggler  zerofill  for ssd's

but what does zerofill  fill drive with  1 or 0's 

does  zerofill fill the blocks with data or  not 

or does it write 1's to blocks  to reset them to clean state 

if the ssd dont support trim  how can garbage collection  know that  data is deleted

[Augeas]

A zero fill would use zeroes from the user's point of view, which is greatly abstracted from the actual values written to flash cells.

If the SSD is quite old (10 years +) then it might not support TRIM, in which case a very infrequent WFS might be beneficial. If nothing is done then there is no way for the SSD controller to know which pages have been deleted by the O/S. When a page deleted by the O/S is subsequently reused the SSD's garbage collector will flag that page as invalid and use a previously erased page. You may have to wait a few micro-seconds in this case.

This is how USB flash drives work as USB doesn't (or didn't until very recently) support SATA protocol.

[julevine]

but if there no trim  enabled then  OS cant    tell the ssd that block is  invalid    so garbage collection will keep moving that invalid data around  

new writes will  go to  all free pages until none are left  or  until it hits a threshold  then garbage collection is forced to look for  invalid data

ive read that  some garbage collection can  somehow  search  for invalid  data  and mark it  without trim telling it  but takes longer 

[julevine]

i guess i dont need to worry about trim  missing   to mark deleted data  because eventually  ssd file system and OS file system of files valid and invalid be be synced  

[Augeas]

That's right, the O/S and the SSD controller will treat the writing of a new file to 'deleted' pages as an update of the pages.

By their nature SSDs run with most of their deleted and TRIMed pages unerased, the SSD is shall we say full, even if the O/S only shows live data. Garbage collection is done nowadays i the foreground as it reduces the number of extraneous writes and also lessens the load on portable devices.

I don't know how the SSD controller can know that pages are no longer used by the O/S's file system. That info is held in the MFT and cluster bit map which the SSD can't read.

[Augeas]

2 minutes ago, julevine said:

i guess i dont need to worry about trim  missing   to mark deleted data  because eventually  ssd file system and OS file system of files valid and invalid be be synced  

I'm not sure what you mean there. If you plough through http://kcall.co.uk/ssd/index.html  it will give me time to go to bed.

[julevine]

thanks for the help