looking for advice

I am going to be setting up a small intranet server. This server is for a small company with 4 users with windows 2000 as their OS. The server is going to be used for a php-mysql database program and for them to save there files to with samba. I am going to be using redhat 7.2 or 7.3 for the server.  

My question is what do you think is the best way to back up the server (money is a issue, not huge but an issue)?

Is it alright just to put in a backup drive and just copy the user files and the database to it.
If so do I just mount the drive backup and unmount the drive when I am done. I read you shouldn't leave the backup drive mounted all the time.

Now one problem is what happens if the motherboard fails they are going to be down until I get a new board. What if I format the hard drive with a fs that can be accessed by windows2000 and linux. So I could pull out the backup drive and put it on a windows machine until I get the server backup.

Or do I just setup the backup drive on a windows machine and backup to it.

Or what about the server with 2 drives that are mirrored to take care of a hard drive failure and a backup drive in one of the windows machine.

What about what kind of backup strategies... Daily, Increment etc...

I know there is no right or wrong answer I am just looking for advice what other people have done before and whats worked and whats not worked.

Thanks for any advice
Who is Participating?
jlevieConnect With a Mentor Commented:
The ideal backup solution is one that provides enough history for the user environment. That almost always means having more than one cycle of backups on hand, with a cycle being at least one full and possible incrementals. How much history one should keep and what a cycle looks like depends on what the user environment looks like. If the user data is fairly volatile, meaning that most of it changes frequently and/or it is of value for only a limited amount of time, then two or three cycles of history maight be sufficient. You'd want at least two to guard against a fault in a single backup or a missing/corrupt file not being noticed immediately. On the other hand if the data is of value for a long period of time then months or years of history would be indicated.

In a like manner you want your backup cycles stored on more than one piece of media. If all of your backup history is on one disk and that fails or gets corrupted you lose everything.

In determining a backup strategy the first thing to determine is how far back in time one might reasonably expect to need to go to retrieve data. The next thing is to determine how much data changes and how often. Those answers can suggest a usable cycle and how much history. A typical cycle would consist of a periodic full and incrementals. The size of the user accessible storage and the frequency of backups will usually suggest the media (CDR/CDRW, tape, DLT, etc).
This old question needs to be finalized -- accept an answer, split points, or get a refund.  For information on your options, please click here-> http:/help/closing.jsp#1 
Post your closing recommendations!  No comment means you don't care.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.