Need some recommendations (Tape Out)

[Jaques(JC)]

Hi there,

So one of my current customers has a policy in place for backups to disk for 60 days, then anything >60 days will be on tape, what would you suggest is the best method here for me to configure the tape out jobs to ensure this does happen, some limitations that I am concerned about are:

Only 1 FC connected LT06 drive with simple library
If you start the tape out job, you have to tick the box to pause the main backup job if its backup to tape (this is required) so you run the risk of not running normal backups for the period of the tape out job running until its finished? ( a couple of days depending on size)

Any ideas?

[Jaques(JC)]

In addition to this, let me ask the following as well:

Say you dont tick the box to "pause" the actual backup job, will both tape out jobs and backup job then run at the same time? Does the tape out job also start by backing up the oldest points in the backup chain and then work its way up to the most recent restore point? If that's the case, that means you can run both backup and tape out job together, and once the backup chain starts to replace the oldest restore points based on date, you should already have that sitting on tape if the tape out uses the same sequence?

I am going to assume that is risky and you ultimately want the backup job paused so it can tape out a 1to1 of what you have on disk without issues in between, my problem there as mentioned before is the fact that you now run the risk of not running backup to disk jobs since its paused waiting for tape out to finish.

[veremin]

Before we starting digging into details, may I ask several questions? What type of backup mode does a source job use? Forward incremental with periodic active full backups? Forward forever incremental? Reversed incremental? What value is set as a retention period for source job? 60 restore points? For how long should data be preserved on tapes? Do you need GFS retention scheme on tapes or simple one is enough?

[Jaques(JC)]

forever forward incremental is used
restore points on source job is set to 60 restore points
retention on these tape out jobs are infinite
simple retention scheme is needed

My concern is if the said policy and backup policy states anything bigger than >60 days there is no room for error and you cannot miss a single "day" or restore points out of that 60, so I am wondering what is the optimal configuration here.

[Jaques(JC)]

any feedback?

[Jaques(JC)]

Am I ever gonna see any form of feedback regarding this post?

[veremin]

Currently there is no policy-driven feature that might give you exactly what you're looking for. However, based on my understand, running a backup to tape job once in two month should answer your requirements. Thanks.