[Info-vax] File Systems

[mcleanjoh at gmail.com]

On Friday, March 6, 2015 at 4:30:07 AM UTC+11, David Froble wrote:
> Stephen Hoffman wrote:
> 
> > Hopefully VSI is aimed at something equal to and preferably variously 
> > better than ZFS, with native UTF-8, and not just a compatible descendent 
> > of ODS-5 here.  But then VSI has a very full project list, too.
> 
> Ok, did some homework.  ( I hate it when you assign me homework )
> 
> For a long time I "knew what I thought I needed to know, and didn't want 
> to be annoyed with spurious stuff".
> 
> Anyway I read some about ZFS.  Interesting.  It sure seems much more 
> robust with regard to data protection and recovery than my much loved 
> ODS-2.  Now I have more to think about.  (I hate when that happens)
> 
> But there is this lingering distrust of "magic".  Work is work, and the 
> more that needs to be done, all else being equal, there is usually only 
> so many work units available and if used for overhead, less is available 
> for what you want to do.
> 
> Actually, I found myself getting a bit excited.  The little bit I read 
> indicates that if anything like this is implemented, some of the current 
> worries will become much less.
> 
> So, Ok, we still don't know what VSI is considering.  But if it's 
> something with the ZFS concepts as a starting point, perhaps we have 
> more than VMS on x86 to look forward to.

ZFS seems to have a lot of overheads.

If I read the information correctly, change one file and a new checksum for that block (or those blocks plural) has to be calculated and written, and a new checksum of the block that contains checksums ... all the way up to the single checksum at the top of the tree of checksums.

Maybe the calculation of each checksum is that messy (done in hardware?) but there's all the disk I/O's that need to be done.  The alternative, of keeping some checksums in memory and only occasionally writing to disk doesn't seem that smart. 

John