first, you really shouldn't be using colon's in filenames.. sure it works fine on the linux side, but what happens if you have to backup to a windows box? ya never know whats going to happen.

well, I tested it.. it chokes in s3sync (probably the internal checks to make sure you have a good source and destination)

it works fine in s3cmd.

I did spend a lil bit of time browsing through the source.. ruby seems like a semi decent code to work with but I've never messed with it..  this seems like something you may be able to ask ferrix to implement. but, he may not because that wouldn't keep it cross platform happy.

oh I'm sorry, I did search around a bit.. guess I should have dove deeper

yea it's strange, even the 'manual' approach doesn't work when going into the directories, which I guess is just a placeholder..

I did find a win based program, Bucket-Explorer, that you can switch between viewing in 'folder view' or full path view.. doing full path view I can see the mysterious keys/files (sometimes only by entering the absolute path), but as for being in 'folder view' even entering the path manually doesn't jive.

and if there is anything I can do to try to finger the culprit on this, just yell..

sucks talking to myself, but at least I'm updating

ok, files seem to be transfering fine.. strange that I can't see them in any other client other than using s3cmd list. is there a way to get the listing of the ACL of the files from s3cmd? maybe somehow some files have different permissions setup?

I was really hoping I would be able to use s3fox or the like to just grab a file here or there when needed, and not have to goto the server and s3cmd when needed.

ok.. for sh_ts and giggles, I went ahead and tried to do a restore with s3sync, and it created all the files that 's3fox' doesn't show.. verrrry strange..  is there another windows based program that I could use to 'double' check s3, or any info that I should know why s3fox is blind to some files but not others?

Strange problem here. I have just recently moved thousands of files to our backup server, and I have our backup server using s3sync to amazon. Log file shows that the files were created (eg. Create node A-E/ACME_DYNAMICS/ACM002C.M09) but when I check from s3fox, the files aren't there.

this is the script I'm running..

Code:

#!/bin/bash

BUCKETPATH=backup-data:/toolcrib/
BACKUPPATH=/backup_data/toolcrib/
EXCLUDE=

date
echo start toolcrib bu...
cd /root/s3sync && /usr/bin/ruby s3sync.rb -rvs ${BACKUPPATH} ${BUCKETPATH}
echo done toolcrib bu...
date

If I were to run it right now in debug mode, here is an clip from debug output.

Code:

local item /backup_data/toolcrib/A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d
local node object init. Name:A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d Path:/backup_data/toolcrib/A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d Size:20176 Tag:4495f29583f73507fc0897f68fb33bce
prefix found: /A-E/ALABAMA_METALS/
s3TreeRecurse backup-data /toolcrib /A-E/ALABAMA_METALS/
Trying command list_bucket backup-data max-keys 200 prefix /toolcrib/A-E/ALABAMA_METALS/ delimiter / with 100 retries left
Response code: 200
S3 item /toolcrib/A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d
s3 node object init. Name:A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d Path:/toolcrib/A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d Size:20176 Tag:4495f29583f73507fc0897f68fb33bce
source: A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d
dest: A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d
Node A-E/ALABAMA_METALS/24-135 WEAR TRACK SPLICE JOINTS.3d unchanged

Going to my backup server in the 'ALABAMA_METALS' directory, there are 125 files, s3sync shows that those 125 files have been uploaded and unchanged from previous syncs, but s3fox doesn't show anything being in there 

any ideas?