dsill@nswc-oas.arpa (Dave Sill) (07/23/88)
Previously, I wrote: > Subject: File checker > Is there a free utility that reads a file containing a list of files > and their modes/owners/groups, compares them to the actual files, and > either reports the inconsistencies or corrects them? Just thought I'd > check before I re-invent. Following is a summary of the replies I've received to date. 1) Two people suggested short awk/ls/diff-based scripts. For my purposes, this is inadequate. I need to check about 1500 files and any form of shell script would just be too slow. 2) Three people mentioned that there is a nice script in "UNIX Security" by Kochan and Wood that does this. Again, too slow for me. 3) One person suggested that such a utility would be useful if it was quick enough. 4) One person suggested trying `cfs' by Rex Sanders. I was unable to reach him. 5) One person suggested that the PD stat program might be useful. True, but not a complete solution. 6) One person volunteered his own unfinished program to do this (Wade Stebbings, wade@violet.berkeley.edu). I would have taken him up on this, had I not already received number 7... 7) Dave Curry gave me a copy of his program `fcc', File Consistency Checker. This turned out to be everything I wanted. It consists of two programs: fcc and fci. Fci takes a list of files and produces a data file for fcc based on the current stats of each file. Fcc compares the data file against the current file system and, depending on how you call it and how you have your data file set up, reports differences in mode, owner, group, file type, link count, size, and modification time. An option to fcc causes it to try to correct owner, group, and mode inconsistencies. On my unloaded tahoe it takes 30 seconds to check 1500+ files. On a lightly loaded 780 it takes a minute to check 1300+ files. I've made several changes to fcc and fci, and have Dave's permission to distribute copies of the source, which includes a makefile and man pages, to anybody that wants them. Drop me a line and I'll send you the shar file. Thanks to all who replied. -Dave
henry@utzoo.uucp (Henry Spencer) (07/24/88)
In article <16608@brl-adm.ARPA> dsill@nswc-oas.arpa (Dave Sill) writes: >1) Two people suggested short awk/ls/diff-based scripts. For my >purposes, this is inadequate. I need to check about 1500 files and >any form of shell script would just be too slow. Did you try it? It ain't necessarily so. Ignoring the annoying problem of argument lists being too long, this should be ls -... lotsafiles | awk '...' | diff - whatitoughttobe which ought to be amply fast unless you're running this thing every five minutes. Kind of academic since you do have a C solution, but automatically dismissing shell-and-filters solutions as "too slow" is a sign of ignorance. Well-done shell programs are often fast enough. -- MSDOS is not dead, it just | Henry Spencer at U of Toronto Zoology smells that way. | uunet!mnetor!utzoo!henry henry@zoo.toronto.edu
duane@cg-atla.UUCP (Andrew Duane X5993) (07/26/88)
In article <8807221517.aa00760@SEM.BRL.ARPA>, dsill@arpa.nswc-oas (Dave Sill) writes: > 7) Dave Curry gave me a copy of his program `fcc', File Consistency > Checker. This turned out to be everything I wanted. > > I've made several changes to fcc and fci, and have Dave's permission > to distribute copies of the source, which includes a makefile and man > pages, to anybody that wants them. Drop me a line and I'll send you > the shar file. > -Dave Dave: Sorry for the posting, but my mailer can't seem to find you.... I am interested in the fcc/fci pair. It would be valuable for many things, primarily for installation control on our machines. Please mail me the shar file at one of the addresses in my .signature below. (decvax or ulowell are usually the best routes). Andrew L. Duane (JOT-7) w:(508)-658-5600 X5993 h:(603)-434-7934 Compugraphic Corp. decvax!cg-atla!duane 200 Ballardvale St. ulowell/ \laidback Wilmington, Mass. 01887 cbosgd!ima/ \cgeuro Mail Stop 200II-3-5S ism780c/ \wizvax Only my cat shares my opinions, and she's out pestering mice.
les@chinet.chi.il.us (Leslie Mikesell) (07/28/88)
In article <8807221517.aa00760@SEM.BRL.ARPA> @cam-cl.UUCP writes: >> Subject: File checker >> Is there a free utility that reads a file containing a list of files >> and their modes/owners/groups, compares them to the actual files, and >> either reports the inconsistencies or corrects them? Just thought I'd >> check before I re-invent. > I need two similar programs: 1) something that will take a list of directories and list each with the mod time of the most recently changed regular file underneath it. It would be nice if it also compared these to a list stored in a file, but that part would be easy to add. This would be a quick check to see if any archive copies are out of date. I tend to compress/uncompress files and move them around too much to tell anything by directory dates. 2) something that will give the effect of "find dir -newer file -print", but would catch files where the creation time is newer but the last modification time is not. I realize that is an odd condition, but the AT&T starlan DOS server sets the mod time according to the PC clock which is often not correct causing files to be missed in an incremental backup based on find -newer. The inode change time and last access time are set from the unix clock, though, so reporting any files where the inode change or the file modification time was later than the comparison file should work. A similar problem exists across RFS mounts also, but there is no easy fix for that, other than keeping the clocks in sync, since all three times are affected. Les Mikesell