Scheduling with Synthetic and Active Full's

[dannythake]

I'm hoping someone can clear up some confusion regarding scheduling and how best to achieve our retention goals.

We recently purchased Veeam 7 and configured it to run 4 incremental's Mon-Thurs and an Active Full on the Friday. Our retention was set at 10, with the goal of having 2 weeks worth of restore points. This worked well but we ran into space issues quite quickly and we would like to utilise to Synthetic option with Rollbacks to achieve a greater retention period.

We have 9 Servers being backed up, 8 of these are tiny and cause no problems, however one of them is 2TB after compression and with only 7TB available disk space its easy to see where are quickly running into issues. We would like to get as many restore points as possible (preferably 4 weeks, or 20 restore points), with a Full backup at least fortnightly (or weekly if possible). We're thinking that Synthetic Fulls with Rollbacks are the best option, but to fall in with best practice we would also like an Active Full once a month. Daily change rate of this VM is between 10-15 GB on average.

How Best would we achieve this, and is it the best solution?

Can you still recover files that have been changed, from a .vrb providing you still have the .vbk?

How does this method change the handling of files that have been deleted previously? can they be recovered?

Thanks in advance for any guidance you can give me,

Danny

[foggy]

How Best would we achieve this, and is it the best solution?

Looks like your storage is capable to manage this strategy, having maximum 3 fulls on some days of the month.

Can you still recover files that have been changed, from a .vrb providing you still have the .vbk?

To recover from a particular restore point, you need the whole backup chain up to the corresponding full backup (VBK).

How does this method change the handling of files that have been deleted previously? can they be recovered?

Ability to restore from some restore point back in time does not depend on the backup method. If you have the whole valid backup chain, you will be able to restore any file contained in the VM at that time.

[dannythake]

Thanks for answering the question. I understand that the chain still needs to exist (including VBK's), I just wanted confirmation that a file-based restored can still be done from the roll-back points much like you would do with an incremental.

Can I ask an unrelated question - If I edit the job schedule for existing jobs, will this cause any issues? It wont try and do a full backup for any reason will it?

I've tried to bounce these question through a support ticket as we have had some problems this week with odd tape backups, but unfortunately I haven't had much response yet

Thanks again

[veremin]

Also, if underlying repository can handle additional workload, you might want to put into use reversed incremental mode, instead; since, it’s the most space-friendly method. Thanks.

[foggy]

I just wanted confirmation that a file-based restored can still be done from the roll-back points much like you would do with an incremental.

Yes, it can.

If I edit the job schedule for existing jobs, will this cause any issues? It wont try and do a full backup for any reason will it?

No, unless you set the next scheduled time for the active/synthetic full day.

[dannythake]

Also, if underlying repository can handle additional workload, you might want to put into use reversed incremental mode, instead; since, it’s the most space-friendly method. Thanks.

Unfortunately I have to backup to tape following these jobs, so that wouldn't be practical in this instance

[dannythake]

Yes, it can.
No, unless you set the next scheduled time for the active/synthetic full day.

Excellent, thank you for confirming these for me.

Would it be possible to ask someone to follow up on my support request 00449176 - I have tapes writing 10gb of data, ejecting the tape and then writing the next 200 gb of data to the next tape - its been happening since Saturday (although only reported to you on Monday) and we're running out of tapes fast!

Thanks

[veremin]

Unfortunately I have to backup to tape following these jobs, so that wouldn't be practical in this instance

Then, indeed, forward incremental is the only way to go.

Would it be possible to ask someone to follow up on my support request 00449176 - I have tapes writing 10gb of data, ejecting the tape and then writing the next 200 gb of data to the next tape - its been happening since Saturday (although only reported to you on Monday) and we're running out of tapes fast!

It’s recommended to create separate topics for different issues, so that, existing thread isn’t overfilled with irrelevant discussions.

So, kindly, create a new topic at corresponding “Tape” sub-forum and provide your ticket number there; this way, we will be able to follow the case and take a more closer look on what is, actually, going wrong with your environment.

Thanks.

[dannythake]

Then, indeed, forward incremental is the only way to go.
It’s recommended to create separate topics for different issues, so that, existing thread isn’t overfilled with irrelevant discussions.

So, kindly, create a new topic at corresponding “Tape” sub-forum and provide your ticket number there; this way, we will be able to follow the case and take a more closer look on what is, actually, going wrong with your environment.

Thanks.

No problem, thanks very much