Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

i have a HP DL380 server, its have 6 harddisk, @72Gb each, i am planning to use centos + cpanel

please suggest how to set daily full backup/daily clone harddisk?

so if 1 main harddisk is broken, i got the backup/clone in the backup harddisk, just need to use the cloned harddisk, or restore it to new harddisk.

its ok if just need to use 2 harddisk, the other 4 is not used, or used for other things like for tmp partition, usr, or other

or if there is better way to backup, i am prefer internal backup in another harddisk in same server, instead external backup that can add cost

share|improve this question

closed as off-topic by ewwhite, Falcon Momot, Jenny D, Rex, mdpc Jun 11 at 16:06

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "Questions must demonstrate a minimal understanding of the problem being solved. Try including attempted solutions, why they didn't work, and the expected results. See How can I ask better questions on Server Fault? for further guidance." – ewwhite, Falcon Momot, Jenny D, Rex
If this question can be reworded to fit the rules in the help center, please edit the question.

2  
i am prefer internal backup in another harddisk in same server -- so you prefer your backups to not actually be backups then? –  RobM Jun 8 at 12:52
    
What you are describing as your need sounds more like RAID than backups. Backups being preferably an external copy of your data on some other media somewhere.. –  Matthew Ife Jun 10 at 20:55

2 Answers 2

RAID is a much better use of these disks. You could setup 4 or all 6 of the disks in RAID10 and get better performance and redundancy - if one drive fails, you don't need to restore from backups, you just need to replace the disk. The server and everything on it keeps running the whole time.

However, RAID is not a backup. It won't help if you delete a file. It won't help if multiple drives fail at the same time, or before you finish replacing the first failure.

Backups cost money. Losing all your data usually costs more money. Especially if this server runs services for clients - if you lose all your client data, you will probably lose all your clients as well.

Backups need to be stored separate from the server. The further away the better. Backups in the same machine can be destroyed by anything going wrong with that server - a faulty power supply frying the disks, a drive controller card going haywire and corrupting everything, etc.

Backups stored in the same room or building are better, but can be destroyed by flood, fire, etc.

Backups stored in an off site location are best - both your server and the site with the backups need to be damaged for you to lose your data. For really super critical data, you might even want multiple offsite backups in different locations.

You also need to test your backups periodically - to ensure they work, and ensure you know how to restore them. I'm assuming you don't want to stop your production server for a day to test those backups, so you need them off that machine and onto a test one.

If you think backups cost too much, consider these points:

  • When your drives die, what will it cost you to try to recover the data from the failed disks?
  • When that fails, what will it cost you in time, effort and money to try to rebuild the server from scratch?
  • What will it cost your customers or your business having to deal with the server being down for an extended period of time
  • What will it cost your customers or your business to recreate all the lost data? Is it even possible to recreate the lost data?
  • What will it cost you in terms of reputation - how many customers will you lose over this?

Those costs all add up very quickly. Backups are a deal in comparison.

share|improve this answer
    
@Gramt wow thank you for the very detailed reply, i am really helped, GBU. btw yes i now i will get external backup too, but i want internal backup too, so i will try raid 10 with 4 disks, so if one drive fails, i just need to replace the disk, with server turn on or need to turn off first? –  Earl Jun 8 at 22:41
    
@Earl with RAID (and hotswappable disks), you don't even need to turn the server off. Pull out the failed disk, insert the new one. That's how you maintain uptime for a simple disk failure. Basically, the RAID controller keeps extra copies of your data spread out across the disks so even if one dies, all the data is still there. For RAID1 and 10, its actually a full extra copy of all data. For RAID5, it's just enough information to figure out what the lost data was and recover it. –  Grant Jun 8 at 22:51
    
how to create hotswappable disks actually? is it already created automatically in raid 10 with 4 disks, or i need to create in the rest 2 disks? –  Earl Jun 9 at 0:21
    
@Earl you really should read your server's manual. You don't create hotswappable disks - they either are, or aren't. If they are hotswappable, the drives will usually be in a carrier that can be pulled out the front of the case without opening the server. –  Grant Jun 9 at 2:08

You absolutely, never ever want a backup to disks in your system. This isn't a backup. If your system dies, it can knock out all hard disks (same with fire, flood and other catastrophes). If your server get attacked, the attacker can delete all disks in the system, etc.etc.

Always back up to external media, preferably off-site or move them off-site in regular intervals.

The cost of doing this is an integral and important part of your business expenses and if you try to go cheap with this, you will regret it.

You can use the additional disks to build RAID arrays that will avoid lengthy downtimes due to broken disks (and/or improve disk performance), but again, a RAID is no backup.

share|improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.