I have a fairly large node -- 6000 items, 6Gb (only 4 are used). Hey, so, out of curiosity I ran a couple of scripts over my store_port file, and I was actually a bit surprised by the results.
What I had expected to see was an overwhelmingly large number of small files -- redirects, DBRs, Mapfiles, and the like. In theory, I think that for an active site and an active Freenet, these files are too heavily favored and will sit in nodes forever. STRANGELY, therefore, I was amazed to see the following results. Of the 6000 files in my node, here's the size breakdown (kind of on a log-ish scale): File Size Count Percentage ------------------------------------------- 0-99 1990 33% 100-999 631 10% 1000-9999 968 16% 100000-99999 999 16% 100000-999999 842 14% >= 1,000,000 516 8% I guess it seems to me that, although the smallest size of files is pretty dominant, as expected, it's not UBER dominant. Now, again, my node is non-standard size -- a lot larger than standard, and a lot more items allowed. So, I wouldn't put it up as "the" snapshot of Freenet. But what I guess I'm wondering is if perhaps we're putting too much blame on small files for reliability. Could there be other reasons for problems with file reliability? ~Mr. Bad -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Mr. Bad <mr.bad at pigdog.org> | Pigdog Journal | http://pigdog.org/ freenet:MSK at SSK@u1AntQcZ81Y4c2tJKd1M87cZvPoQAge/pigdog+journal// "Statements like this give the impression that this article was
