Hi Guy,

Thanks for that. Especially neat about the physical layout versus logical.

I'm 95% done testing the phenomena and will be reporting my observations in a new thread later today (hopefully).

What I can see by now is that continuous is usually best.

cheers


>Hi Jim,
>
>Your question looks like you just want some affirmation that file fragmentation is ...still... bad.
>
>You're right that disks, processors, etc., are faster these days than what we may have worked with in the past. That means that as a newer disk gets fragmented, we may not see the same degree of performance loss.
>
>However, the principles have not changed. If you have a busy system with large tables and you have excessive file fragmentation on that disk, you will almost certainly have slower performance than the equivalent scenario with minor fragmentation.
>
>The difference in performance will depend on sevaral factors: How many users? How busy is the drive? How big are the records? How much data is being read or written from your application at different times? etc....
>
>You can generally expect that newer equipment is faster and therefore may make fragmentation somewhat trivial for a given application. But this cannot be taken as a rule across-the-board for all applications.
>
>As a small side note, a while back I was reading about fragmentation and was surprised to learn that a non-fragmented file is NOT sitting on the disk in contiguous sectors. When a disk is formatted, its sectors are laid down in a non-contiguous pattern. This is to accomodate the delay between reading or writing buffers (for a sector) and the disk drive preparing for the next sector. Since some small amount of time passes and the disk drive has kept spinning, the head is no longer above the next contiguous sector when it is ready for its next read/write. To account for this and to help get maximum efficiency from the drive, sectors are "interleaved" in a non-contiguous fashion. So instead of seeing sequential sectors like 1,2,3,4,5,6, etc, they might be ordered on a disk drive as 1,4,2,5,3,6 etc.
>
>So when we speak of fragmentation on the disk, we are refering to "logical" contiguous sectors; as opposed to "literal". As I understand it, every file on the disk (larger than a sector) is literally non-contiguous.
>
>For us application programmers who aren't engineers that have to deal with things at this level it's just plain easier stay at the conceptual level and speak of contiguous files in the logical sense. (I just thought this information might be interesting to you.)
>
>Guy
>
>
>
>
>>I tried this in the CHATTER forum, but an absence of response prompts me to re-try here.
>>
>>Keeping in mind:
>>1) Modern Windows systems are multi-tasking systems.
>>2) Windows itself (and its components, like IE) make significant 'quiet' use of your HD space for all manner of files, large and small.
>>3) Other applications (MS Word for example) can use HD space 'quietly' too.
>>4) Modern HDs are fast, processors are faster yet, and RAM is plentiful.
>>... what hard facts are there to back up the axiom (it is essentially an axiom today) that fragmentation is bad?
>>
>>That fragmentation is bad is so prevalent a concept that I must be missing something obvious. What is it?
>>
>>Thanks for any/all input on this issue.
>>Jim Nelson