Operating System - HP-UX
1855666 Members
10628 Online
104112 Solutions
New Discussion

10.20 Disaster Recovery Strategy

 
Brett Hamilton
Occasional Contributor

10.20 Disaster Recovery Strategy

We use an older 700 series Workstation running 10.20 that has became a critical system in our operation. The system software is frozen at it’s current level (because of the applications). I currently do a “make_recovery” every few months and we do nightly backups of all the user accounts. I have never had to use the recovery tape and am a little nervous that things might not go correctly. My first question, is it possible to have an identical system on the network and every night all files mirrored to the second system, so change over could be very fast? The original system has two disk drives installed. Any advice will be apprec
3 REPLIES 3
KapilRaj
Honored Contributor

Re: 10.20 Disaster Recovery Strategy

I would ...

1. Run a make_recovery every week
2. run a backup every day

Using the above i would try restoring the whole applications on another idntical system every month or so. I would call it recovery test.

when u do the recovery testing , please shutdown the active one. ;)

Regds,

Kaps
Nothing is impossible
Sridhar Bhaskarla
Honored Contributor

Re: 10.20 Disaster Recovery Strategy

Hi Brett,

Few precautions.

1. If possible, have two more disks attached to the system and mirror your root disk and the data disk. That will prevent disk issues.

2. Continue to take 'make_recovery' tapes. Keep one set of tapes offsite.

3. Also take 'fbackup' of the system everyday or once in two days. With fbackup, you can schedule 'incremental' backups so you don't have to take full backup the system everyday, But it will let you recover completely (upto when the last backup was taken) in case of issue. If there are any user errors and if you need to recover, you can do it easily from fbackup.

4. You can have an identical system. What I would do is to use the 'make_recovery' tape to clone the original system and change the network configuration (/etc/hosts and /etc/rc.config.d/netconf files). So, in case of system failure, you can replace the system (assuming that there are no softwares that are CPU id licensed). Mirror/UX will take care of disk failures.

-Sri

You may be disappointed if you fail, but you are doomed if you don't try
kenny chia
Regular Advisor

Re: 10.20 Disaster Recovery Strategy

My first question, is it possible to have an identical system on the network and every night all files mirrored to the second system, so change over could be very fast? The original system has two disk drives installed

-> Try using Remote Sync. Remote Sync can be easily setup to copy differences from one server to another server. This allows quick update to the second system since only differences are copied. You can get it at http://rsync.samba.org/

Example usage to mirror from first system to second system

/usr/contrib/bin/rsync -av --delete root@first_system:/from_dir /second_system_dir



All Your Bases Are Belong To Us!