lotsa memory usage during file itemization

[fgagnon]

My test case started with a 50000 file folder, & later 40000 (see my comcast folder for the matlab script I used to generate).
It seemed like memory issues should have been nil since total use was well below amount installed (less than half).
It was remarkable that there appeared to be break exactly at a power of two (2^13).
First try was to move 10000 items out to a second folder, to leave 40000 in the original (because generating the directory fill slowed notably after 40000). That move produced a 'hang' after about 6K; Thinking binary, I explored sizes at 2^13 boundary as a suspicious count-point -- but from your statements, that may have been coincidental.
I'm not particularly motivated to continue this investigation because it's such an extreme case; but I thought it was interesting enough to report the 2^13 "coincidence" as being a potential indicator of what to look for if anybody were really concerned.

[nikos]

obviously i am concerned but my test this morning used 10000 files and had no problem -- but i was deleting and you was moving files

[johncesta]

> i make no pretences, x2 was designed with speed, not memory, in mind.

Mine wasn't a test. It was a practical application that prohijbited the use of xplorer2. The way I described was pretty much the way it happened. My server is a 2.4 gig processor with 1 gig of ram. It runs 24 x 7 and has zero other problems.

If that's the way xplorer2 handles that many files then that's that. It neither handled them with speed nor was it memory efficient. My Winbatch program (however it is consrtucted) did that particular job faster and with less memory. That's what I'll use next time I encounter a similiar situation.

John

[nikos]

x2 was designed with speed, not memory, in mind

let me rephrase that to make it clearer:
x2 was optimized for accessing small to medium size folders, which is what most people do most of the time

[fgagnon]

WARNING...
For the time being, I advise against anybody trying to intentionally repeat my test case.
This morning, I could not 'boot' from that drive [windows2000, sp4].
I believe the problem to be not related* to x2 handling files during the experiment; but until I can verify what happened yesterday to keep NTLDR.exe from running this morning, it would be prudent to not try what I did.
PS - First reports from my I.S. dept indicate the disk is ok & files are intact; but I want to get a full report from them and check a few things out to try & establish what happened.

* I think I did something careless ... but I'm not sure.

[edit, July 28]I.S. dept reports Master Boot Record corrupted, but they couldn't repair it without 'rebuilding' the drive from scratch. We will never know whether it was a random failure, or something I did, much less whether it was related to trying to manipulate 0000's of files at once with x2. [/edit]