On Saturday, January 24, 2015 at 1:29:28 PM UTC+11, Nikolaus Rath wrote:
>
> > 
> > Some applications (notably PostgreSQL) do not update inode dates when 
> they 
> > update files, specifically to reduce IO load. ie. the data is changed, 
> but 
> > the modification dates (and quite probably size) are not. 
>
> Are you sure that's correct? 



No. It's something I vaguely recall from years ago, and it may not relate 
to unix, and it may only relate to flushing vs update dates, which would be 
irrelevant in the s3sql context.

 

> > So the -c option becomes important, at least in this case. 
> > 
> > I also (sometimes) change the file modification dates on photos to the 
> > original photo date after trivial edits: eg changing EXIF data. In this 
> > case the date and size remain the same. 
>
> That might cause problems indeed. But seriously, I'd simply not do that 
> instead of trying to patch S3QL to support a bigger cache. 
>

While it's not a sufficient reason to patch, the 100% cache does have other 
advantages (like allowing the file system to work when the network is 
down/disconnected).

 

> The checksum is calculated before compression.


Ah, excellent! I did wonder about that, but mistakenly believed the cache 
was compressed.

 

> > 
> > Normal Copy: 
> > 
>
> No. It's 
>
> 1. Calculate checksum 
> 2. If different: 
>   a. compress 
>   b. encrypt 
>   c. upload 
>
> > rsync -c copy with complete cache: 
> > 
>
 

> No, that wrong as well. The cache is stored uncompressed. 
>

So it becomes:

1. compare local data 
2. if different: 
  a. checksum
  b. compress 
  c. encrypt 
  d. upload 

which is of negligible value IFF file size and date are trustworthy (which 
still concerns me).

I must admit I see the 'rsync -c' option as a much stronger guarantee than 
a copy based on file dates & sizes, and this could be irrational. But, if 
100% cache can be achieved at nearly zero performance impact *and* it 
supports rsync -c as well as an offline mode, this seems to me like a big 
gain at minimal cost.

 

-- 
You received this message because you are subscribed to the Google Groups 
"s3ql" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to