Carlos E. R. wrote:
The Wednesday 2005-01-12 at 14:46 -0500, Jeffrey Laramie wrote:
I use a cron script that runs at night which pipes the results of mkisofs into cdrecord. This works fine as long as the files don't change during the burning process. Something like this:
mkisofs -v -R /data | cdrecord -v fs=12 speed=4 dev=ATA:1,1,0 -
If your a competent script writer (I'm not) you could improve on this by creating the iso first, burn it, then delete the iso image. That would nearly eliminate any chance of files changing.
With an intermediate step, you can also compress the image. The linux kernel can later mount the compressed cdrom or dvd transparently. I don't know if any gui can do that.
Really? That's interesting. How to do this? I would expect that this would work better than this: I made a compressed tarball of my stuff, about 500MB after compressing, 950MB before. I copied the .tgz to a DVDRAM. Then I clicked the .tgz in Konqueror, which uses whatever to open it and display it just like an ordinary directory, which of course it is not. This is VERY convenient for small archives, but for a 500MB .tgz on a slow medium, it is terrible. It can take literally 15 minutes to access a single file. If I do not compress the tarball, and just copy the .tar to the DVDRAM, then it works much better, and is actually a fair solution, since it also optimizes the write speed to the DVDRAM. Writing a single large file (of whatever size) is fastest, compared to writing many small files, since that requires the drive to seek seek seek all the time (to write the directory entries and data, which are in different physical locations.) Since I have only about 0.9-1.0GB of daily data to backup, I don't mind using an uncompressed .tar on the DVDRAM. The disk is still 4x larger than my backup size, so it will be many years before my home dir expands to be too much for the DVDRAM. So I think what you are talking about is different. The .tgz is not a *filesystem* proper, while a compressed ISO is. So I would presume that this could work much faster than my attempt at accessing a 500MB .tgz. Good day! -- ____________________________________ Christopher R. Carlen Principal Laser/Optical Technologist Sandia National Laboratories CA USA crcarle@sandia.gov