Hello,
I'm looking for a software that backup data folders:
- The master folder (A) should be untouched
- The destination folder (B) should be the perfect copy of (A)
- Case sensitive
- A file with the same size and date but with different metadata should be updated on the (B) folder
Exists what I looking for?
Yes Robocopy, you can use it from the command lline
robocopy A B /mir
Creates a perfect mirror of A at B without copying files from A that already exist on B....
Hi there
If you have a NAS system on which you can install software then install GRSYNC on it -- that does the above and a whole lot more with a GUI if you don't like the command line - it's based on a very useful command rsync -- also having a backup AWAY from the original machine IMO is always a lot better than storing it online.
You can use rsync as well on Windows if you install the WSL. quite a good reason alone to have the RSL !!!!!.
E.g apart from identical copies -- you can get lists of what's in A and not in B, what's in B and not in A, text searches (that work) on directory / file name, update target (copied file) only if newer or whatever. Also you can use it the OTHER way around --to restore the backup file(s) to the original destination / source disk. Also has the ability for sub directories and directory re-cursion. Incredibly useful piece of software.
Robo copy works but if you DO have a NAS where you can install software on it --why not take advantage of the facilities it offers.
Cheers
jimbo
I'm sad: Robocopy doesn't is "case-sensitive" and doesn't copy changed metadata if file has the same size/datestamp.
So it's not good for both my needs
I have some mp3s, I change their metadata (i.e. I change "YEAR"), but I leave the same timestamp and the size is the same (even if I change metadata the size of the MP3 result the SAME).
So the destination isn't mirrored correcly.
For the "case-sensitive" issue, another very big problem for me.
I use a commercial backup system, Aomei Backupper Professional, but lot's of alternatives exist so I'm not pushing this specifically. I have large(ish) collections of Video [300+], Music 42000+ and image [35000+] files, and these too are regularly changed at the metadata level, even to the level that, for example, the MP3 music files will have a data update when accessed/played.
I use specialist programs to organise my files rather than rely on windows limited file manager, limited with regards to metadata management, (it's a good product for it's designed purpose), I use MediaMonkey for Music and Video file management., and I use Lightroom for my Images and RAW development.
I obviously need to back these items up on a regular basis, I did as you did and looked at monitoring changes for splitting off for intermediate backups but found that even when the software had enough control to split things off the time taken for the backup run increased dramatically.
In the end I just decided to run only full backups, data storage is relatively cheap these days, My time is less so.
If I need to organise my files in a particular way based on their metadata I do so in the specialist data management programs, Yes I also organise by specialist filenames and other simple methods that gives a basic order that is used by Windows File Manager, but the specialist software allows total granular sorting, and in the case of Lightroom it renames all my images into a predefined file template automatically during import. Mediamonkey, does the same for my Music/Video files even to the extent of creating artist & album folders, and the filenames from the internal metadata. Lightroom backs up it's own database on closing the application, I do this weekly but can be done more often to suit the amount of change happening, these backups are included in the regular data backup routine, (my image files are mostly photographs that I have taken or artwork built from them so are not recoverable if a total loss occurred )
The "Horses for Courses" principle is still relevant in current computing practice. If you wish to run backups using differential or incremental backup schemes use whatever the backup system allows, use file naming templates either automatic or manual to organise files in Windows explorer, but if you want to specifically maintain and organise data a specialist application is best
Hi folks
@Barman58
I always think you have good suggestions etc - but how do you solve the problem of files etc with paths longer than 266 Chars just using pure Windows tools -- I've a load of stuff on a 20 TB RAID 0 array with some really long names etc. RAID 0 works for me as I regularly backup new / changed files to external HDD's just in case -- had actually NO HDD problems in years though with RAID 0 -- fast I/O and full use of all HDD's (4 in this array).
This will work in CLI mode using WSL on Windows (rsync) -- or on a GUI running any sort of Linux as a VM or on a NAS.
There's zillions of options on this -- screenshot here just shows basic options
Also doesn't suffer from the 266 char folder/file etc that Windows has -- (internally that restriction on Windows I know has been removed but about 99% of programs manipulating File things on Windows are still based on the old API used by file explorer etc so the restriction still applies).
Windows display (including Robocopy etc etc)
Actual directory :
I have around 4TB of music from various sources and some path lengths are > 266 chars -- file explorer (and things like ROBOCOPY) give up the ghost with long file names and revert back to the old 8.3 format so totally useless in my case.
(And I'm not going through a futile exercise to rename this stuff either).
"Case Solved M'Lud -- Next Case please" !!
Cheers
jimbo
THe file and folder length is only an issue for a minority of naming systems, for most users they will never come close to the limit.
If I want to sort and access a dataset with multiple fields I do not create a single field with everything in a single word - which is the only way that a file manager will work (and badly too)
Using a database with multiple fields is a simple thing to control, monitor, and use, with even a single datatable, multiple tables give virtually unlimited sorting and data management.
If I have a nail to drive into a piece of wood I use a hammer not a hand saw
Rather than try to use a file manager as a DBMS use a DBMS !!
My Photo image files are named with their full date and time of creation down to the sub second - YYYY-MM-DD-HH MM-SS-nn [nn being the shot number in that second]
This works with a similar date based folder structure,
all other data is encoded in the file Metadata.
I can open Lightroom and find any file within seconds based on real relevant data, all pictures of "ducks taken in May in cardiff" no issue at all
Case closed
Hi there
@Barman58
Case NOT quite closed !!!
Good reply BUT : what if you need a CLIENT multi-media player (I use KODI) on a remote piece of equipment --remote TV's etc to PLAY the particular track / film / TV show.
I'm using amazon firetv boxes connected to remote TV's with KODI installed as a client (could be other clients --I just like KODI) -- now the android system will play the relevant track etc (selected via its file manager) and has no problem with the length etc.
I'm not sure how one would use an android client to play the relevant multi-media file on the device using some type of propietary DBMS.
There might be some that work OK but my experience of propietary DBMS's is that they can break and then you are hosed.
Opensource MariaDB and MySql are OK but getting a client media playing application to simply play your track / video on the remote box -- nothing is easier than just simply browsing (on the client Amazon TV stick) to the appropriate file and playing it. Amazon TV boxes connected to TV via HDMI (up to 4K UHD files can be played as well with these amazon boxes and the 55 inch samsung TV's).
(Not trying to give you a hard time here BTW !!!!).
Typical Network map of what I have -- often there's 2 more amazon boxes with 2 more TV's and a load more computers on the LAN!!!
Cheers
jimbo
Quote
