Romex Software Forum

Hello Support,

I would like to know if Fancycache cache read / write session of defragment utilities, such ask Diskeeper, Perfectdisk, and etc, that have auto defrag feature.

I notice, in the perfomance monitoring tool for Fancycache, that there were high read and write cache hit during fragmenting. My worry lead to the wear and tear on the ssd l2.

Is there a possible way for Fancycache to auto stop caching when defrag is taking place and re-able after it stop. Also that would reduce waste resource so the defragmentation can goes faster.

Thank you for the feedback.

Since there are too many defragment tools, currently you may disable/stop caching manually.

We'll check to see if it is possible to do it automatically.

I understood that the Windows Defrag utility (which provides the defragmentation API to thrid-party defrag utilities) bypassed cache and wrote directly to the disk. Does FancyCache sit underneath this, so that the defrag data is buffered through FancyCache?

swmspam wrote:I understood that the Windows Defrag utility (which provides the defragmentation API to thrid-party defrag utilities) bypassed cache and wrote directly to the disk. Does FancyCache sit underneath this, so that the defrag data is buffered through FancyCache?

Yes, since it's a low level block-based cache, it works virtually independently of the file systems,..

This suggests defragmentation with fancycache active is a very risky proposition indeed.

As an engineer with Diskeeper for more than 20+ years I can tell you there are no known compatibility problems between Diskeeper and Fancycache. Diskeeper focuses more on the prevention of fragmentation by helping NTFS select a more appropriate file allocation size such that the file does not become fragmented in the first place. Under this condition, not file movement is necessary.

In addition, we wrote the MoveFile API for Microsoft and the defrag algorithms used by Diskeeper only need to move the files that gain a performance benefit from further defragmentation and/or when the consolidation of free space provides a substantial benefit in performance. Not all files are moved each and every time defragmentation starts up, in fact Diskeeper has over 20 different defragmentation engines, each designed for a specific file system allocation condition.

The bottom line is that performance it improved and the I/O traffic is over the long run is greatly reduced by using Diskeeper.

Hello HowardB,

Thanks for explaining how Diskeepper work and engineer with such sophistication. I use Diskeeper at home and on the servers at work. It had been savior for me many times.

As Fancycache, like any other caching utilities, works at a level underneath Diskeeper, I would presume that it will cache all read and write during defrag. Running a SSD Caching solution, along with auto defrag uitlities like Diskeeper or PerfectDisk, can easily kill a SSD. Fancycache should have a mechanism to detect a defrag in process and pass it directly to the hard drive. I would assume that once Fancycache have the persistent cache in place, the hot data would get push out during defragmenation as data get move around also

I could easily see and feel my Seagate Momentus XT 750 hybrid drive is slower after it was auto defrag by Diskeeper. Rebooting my laptop 3 times return the speed back to almost SSD state. What I am relating is that defrag will push out usable data that is already in cache and the wear - tear on a SSD L2 of Fancycache.

Thank you HowardB,

Your explanation was excellent and helps dispel confusion - a lofty and admirable ambition.

Therefore, it appears that disk I/O is greatly reduced with a Diskeeper/FancyCache combination. Obviously, FancyCache significantly reduces I/O (I regularly see 30% read hit rate and 70% write deferral). When used in conjunction with Readyboost, the system read hit rate is considerably increased. Adding Diskeeper further optimizes the write I/O.

Any comments on CleverCache?

Hello Everyone,

Since Diskeeper doesn’t see the FancyCache as a separate storage device, the reads/writes that do occur from an application (defragmentation activity included) over to NTFS.sys will be processed according to the various storage drivers on the stack. I would agree when the accumulation of fragmentation over time results in a situation where if a defrag is performed, the data blocks stored in cache are subsequently flushed and replaced with more recently accessed data blocks then one could potentially see a short term negative effect on cache hit rates.

To solve this problem Diskeeper engineered a technology called IntelliWrite and that prevents the vast majority of fragmentation from occurring in the first place. By helping Windows (NTFS.sys) allocate a more appropriate amount of data in the first place, data is written contiguously in a more sequential rather than random fashion. This will improve cache hit rates and precludes the need for defragmentation after the fact. On average, IntelliWrite prevents 85% or more of the fragmentation without the need for file movement. As a result there are no wear and tear issues to speak. In fact, erase activity as well as the Write Amplification index on the SSD is greatly reduced and documented with the use of Diskeeper’s technologies.

I haven't played with CleaverCache is such a long time, I don't feel qualified to comment.

I would not suggest to use any other memory manager in addition to superfetch, readyboost, fancycache, and diskeeper combination. That would be an extra layer that is not needed. It will do more harm as Windows 7 is more than capable of caching data and managing ram at the file level.

If you are using Windows XP, then I can suggest that.