Update 4a and slowly moving towards incremental backup

[chrisit]

Hi,

I have read Gostev's email about U4a and upcoming changes. Would like to know what happens if we turn of the Full Active backup and simply use incrementals in a scenario where backups some of the backups are deleted from a repository. Obviously breaking the chain - this is similar to silent data corruption, but the file simply is not there anymore. So how do you tackle these?
One thing we don't want to end up is in a place that we are unable to recover data...
Currently, if the whole chain fails we still can recover from last week's full backup.

Thanks.

[Gostev]

Hello,

In general, this and similar scenarios is the reason behind 3-2-1 rule, which requires you to have a copy of your backups on another storage. So, to answer your question, you just restore from the copy of your backups. Do you copy your backups today? And if not, how are you protecting from the deletion of all backup files (not just some), or the complete loss of the storage behind your primary backup repository?

Also, I think you misunderstood my post in general. I was talking about reliability aspects of forever-incremental backup and no need for active full backups (when all data is read from the source storage in its entirety). I did not discuss the need of periodic fulls in general, which really depends a lot on customer requirements and also backup repository type. For example, with ReFS customers like to do forever-incremental backup with synthetic fulls (instead of active fulls), which will result in the exact same backup chain as you have today - but without having to pull 100% of data from the source storage periodically.

Thanks!

[chrisit]

It's all clear now, thanks for the answer.