Bypass File System filters option

[TomTiddler]

Is it possible to get a little more detail on this? Especially the phrase "on-the-fly data manipulation". On the face of it, it seems possible that almost any database operation might qualify, in which case I'm gonna have to do a whole lot of thinking about what gets placed in the pools. 

Looking for

1) Some estimate of the potential gain from this.

2) Clearer notion of what may experience negative effects.

DrivePool user for years now, and talked at least one buddy into using also. Love this product, and can't say enough good things about the support :-)

[Christopher (Drashna)]

File system filters are used by a lot of different software.

But the main one that you may be using, is that they're used by Antivirus software to scan files that you access, and to outright block access to potentially malicious data.

However, they're used bu a number of encryption solutions to decrypt reads and encrypt writes.  And the Windows Server Deduplication feature uses them to splice together deduplicated data (it stores duplicated data "elsewhere" on the disk, that isn't normally accessible). 

The problem is that many AV solutions use various tricks.... The worst one is that if they see the same file being accessed, instead of handling it normally, it waits on the first request.  Normally, this would be fine, but .... for DrivePool, that would lead to an I/O deadlock, as the second access (the one now "waiting") is from the first access... meaning that the file can NEVER be accessed.   This is (thankfully) fairly rare, but we've seen it happen often enough.... 

 

As for gains, namely performance.  Each filter that has to process a file is one more delay in accessing the file.  This can add up quickly, if you have a lot of filters, or filters that are "slow".   

 

I think that should really cover everything you wanted to know here. But if not, then please do ask for clarification! 

(and glad to hear it!)