Your opinion on Clones

[iamwhitey]

Currently Clones are run for my company by scripts written years ago that I just adopted when i took over the backup system. The scripts are pretty basic and work decent.

Now i am condidering changing clones away from these scripts. We only have 5 customers that get off site clones so its not a ton of work to make the migration.

My idea was to create a seperate backup group for each company and then just change the schedule to run the Full backup on their clone date and thats what would be sent off site. Now the downfall to this is that it will eat more Network bandwidth since it would essentially do 2 fulls per schedule clone day. not a huge concern, but could be in the future. The pros of this setup are I have acknowledgement that the clone ran successfully, I have an easier way to set the retention on the cloned data, I can successfully track the data being cloned, and changes can be easily made by even a novice user.

So I ask, how do you do clones? which ways have proved easier for you? Which method do you think is more efficient and accurate?

thanks

John

 

[605]

Although you may clone betweeen two storage nodes this is rarely done. Clones are usually run between media from different pools on the SAME node.

BTW, as the backup has already been made, a clone is not another backup of the client but just copying the data from one media to another. Unless you have a NW 5 Workgroup or Network edition, you will not use the network at all ... as long you stay at the same node, of course.

 

[iamwhitey]

I am not quite sure what gave you the idea that i was using multiple nodes, but for futher clarification I am using only one node.

here is my logic. So there are scripts on the Legato server that start the clones and tell it what SSID's to clone. A clone is a copy of the full backup from a certain time. Thats fine I understand that will use no network bandwidth.

my idea was to create a seperate Group in Networker to do clones. For instance say thier is a normal backup group named John_Daily well I would make a second backup group named John_Monthly_Clone. John_Daily will backup every day, where the John_Monthly_Clone will backup a Full once a Month.

By creating this second group you are providing a visual for a person to see if the clone was done, the details on the clone and a better way to chart the data usage.

are you understanding what I am saying?

 

[605]

Of course you can create another group to create such backups. Please understand that those will of course be new backups, not clones, in NW terms. That's why i got confused.

Create a group, set the level for the group to full and disable autostart. Then if you need to run it, simply start it manually or set autostart=enabled for a night.

 

[iamwhitey]

yeah i am not asking how to do it, I know how to do it. I want to know people opinions on the differences of cloning processes.

What does everyone else do? What have they found that works better?

 

[RMinNJ]

Sorry, I got confused also. Are you saying you'll run
a second full backup instead of a clone? Or are you saying
you'll simply turn on auto-cloning for the full group that
is now run seperately?

 

[VicMcD]

There are three possible ways to do this


Taking a monthly backup and sending the tapes offsite will be quicker but you have no guarantee that you will be able to restore from the tapes.

Cloning the backup and sending the clones offsite will be slow and again you have no guarantee to be able to recover from the clones.

Cloning the tapes and sending the originals off site is better as you have proven that you can read the data from the tape but you cannot be sure that the data is valid.

What I do with my monthly backups is to run a test restore before sending them off site. This is not perfect as I would rather clone them as well, but cost and time precludes this.