UNIX for Dummies Questions & Answers

does anyone have any idea of what this means and if you do, please elaborate more on what you know about this and the importance it carries along.

by the way, we dealing with Linux and Solaris 8 systems here

When you have a disk mirrored, both copies must be the same. Sometimes, they are not and then you need to synchronize them (usually by running a command to copy one to the other - realize this is not a cp command but is being explained that way)

With Solaris you would be probably using one of two software products - DiskSuite or Veritas. The commands for each are entirely different. Suggest you figure out which you are using and read the manuals.

$ df -kl

If the output has devices that contain /md/ - Disksuite is being used there.
If the output has devices that contain /vx/ - Veritas Volume Manager is being used.

Opinion - DiskSuite for OS drives - Veritas for Data drives. Veritas is a pain in the neck when dealing with OS drives (which is why most don't use it for OS). DiskSuite isn't as robust for Database/Data drives which is why you spend some $$$ and get Veritas. DiskSuite comes with Solaris 2.6 and above. Not sure about below 2.6.

Check out docs.sun.com for on-line docs on DiskSuite.

Check out Veritas Support for on-line docs on Veritas.


For Linux Mirrors check out this link.

For general info on Raid levels

The subject you posted "Run process to resynchronize mirrors during database backups" - was this an error while running backups? If so - is it happening on both Linux & Solaris? What type of backup (data dumps, ufsdumps, vxdumps, 3rd party backup solution {version}). Give more info unless you are just looking for information on Raid.

The thing here is that, I am learning about this mirroring stuff.


I was reading a "Linux Red Hat 6.1 Bible" book and they talked about backing up data to hard drives and they go on to list the advantages and disadvantages

Now, they say that for Linux Red Hat systems, I can download the file mirrordir-0.10.39.tar.gz. after downloading, I then unzip and cd to the directory and type the commands "configure" and "make install" to install the new downloaded file.

now, after all the installing is done and assuming that I already have a ready second hard drive. They say I have to make a new directory by typing "mkdir -p /mirror/home"

assuming I want to mirror the home directory on my original disk to the second hard drive, I have to mount the second hard drive by typing "mount /dev/hdb5 /mirror/home"

/dev/hdb5 stands for the second hard drive.

now, to do the actual mirroring, I have to type "mirrordir /home /mirror/home"

this command causes the contents of the /home directory on my original disk to be exactly duplicated in the /mirror/home directory


now, what I want to know is that, to those who actually do this at work, is this example I got from the linux book a good example for a beginner like me to follow?? and if not, please, set me right.

thanks

The example in the book is a little confusing because few people would have a database in their home directory A more practical example would be a MySQL user who has the database located in /usr/local/mysql/data and is running an application than requires mirroring...... then you might have the mirror on /mirror/mysql/data.