UNIX for Dummies Questions & Answers

Need some advice and guidance for this UNIX beginner. Due to downsizing I have inherited the SysAdmin duties..(sigh). Please excuse and forgive me if I use the wrong terms below....

Situation:

We have UNIX ( Solaris 7/8/9( it varies) on Sun Ultra 10's) servers located at several global locations - in Malaysia, Mexico, and South Carolina. We access these from our New England location via the corporate network, thru firewalls located at these sites.

At each of the sites, we have a UNIX server ( Ultra 10 running Solaris) that acts as a gateway server, that bridges from the corporate network, to a isolated "Test Engineering" network within the remote factory. Hanging off of this "Test Eng Network", are more Ultra-10 Servers, used as NMS Network Management System servers to run Systems level traffic tests on test equipment.

Goal: Is to make duplicate clones of each of these servers hard drives, so if ( and When) they die and crash, we can rapidly power up and attach the cloned server ( with duplicate hard drives) and prevent downtime.

Problem:
1) These Ultra-10's have 2 hard drives in them, attached to the IDE bus connections - thus no way to attach a new fresh drive and do a "DD" duplication, as both IDE connections are in use..

2) Due to the remote location , and lack of expertise in IT capability there, we also can't rely on the local people to dismantle and attach
the fresh drive to do a DD clone anyways - too risky.

3) We want to avoid powering down these servers, as we have had problems in the past of them rebooting without hard drive crashes due to their age.

My Question:

Is there a way to make a copy of each hard drive, by doing a copy over the network?. Would this involve partitioning the fresh drive( as an attachment 2nd drive on a 1 drive Solaris system in my lab) to match the same partitions as the ones in the remote global location that we want to copy , and then copying all of the files in each partition/directory.??

Or is there a way to do the equivalent of a DD cloning operation between 2 systems that are on a global network...
??


Any ideas or advice is appreciated.!!

Ignoring the problem of software recovery for a minute, How do you plan on getting the new hardware onsite and installed once it does fail?

You can certainly use dd across a network, just omit the 'of=' parameter and stream it somewhere (probably over ssh and to a file on a central system).

Flash archives are effective for this sort of thing too.

However, I would think that solstice disk suite (aka solaris volume manager, aka logical volume manager) to mirror between your two internal IDE disks would be a better option. Can you reduce your usage to only one disk? If so, you can mirror to the second disk. That way a disk failure won't even stop the server. Then, at your leasure, when you have a suitable time to have an outage, arrange for Sun (or an engineer you trust) to come onsite with a new disk. One the disk is replaced, the server is booted back up and it's easy to remotely reestablish the mirror again.