Local storage on backup server

[omfk]

Hi all,

are there any guidelines who to use local storage on the backup server as backup targets?

I've set B&R9.5 Essentials on a HP ML310e Gen8 v2 running 2008R2. I would like to have the first copy of the data on this server as well instead of using the backup server only as a distribution machine. From this first instance I've set up backup copy jobs (at the moment separate FreeNAS machine). This workflow works fine so far.

Currently we have one Hyper-V machine with 3 VM and all together 300GB of data and second Hyper-V machine with even less data (100GB). In addition we still have our domain controller with the file service which is still a physical server. This one I would like to backup, too (first with Veeam endpoint and later Veeam Agent for Windows). The file service, however, is approx 3TB of data (daily change marginal [<10GB], most of the time).

How can I include sufficient hard drive capacity in the HP server? It has no hardware raid, but only software raid.

Is using a software raid an "approved" solution?

What about JBOD?

Would be 20012R2 a better solution as underlying system?

Thanks for comments and advices,
Frank

[PTide]

Hi,

Is using a software raid an "approved" solution?

Techically you can use software raid, but I would not recommend doing that for your primary backup storage. It's B120i in your HP, isn't it?

What about JBOD?

JBOD is OK to use if you're fine with the idea of loosing a part or the whole set of your primary backups on a day when one or more harddrives in your JBOD fail. On the other hand if the backup copy job copies every new backup immediately then it might be not so bad as it sounds.

Would be 20012R2 a better solution as underlying system?

Depends on what you're after. This thread is worth checking.

[omfk]

Hi,

Techically you can use software raid, but I would not recommend doing that for your primary backup storage. It's B120i in your HP, isn't it?

It's B120i. It remains the question if the purchase of a hardware raid controller makes sense. Basically I'm looking for apossibility to have a repository which isn't split into individual hard drives, especially for the 3TB data from the file service. Are there other solutions?

JBOD is OK to use if you're fine with the idea of loosing a part or the whole set of your primary backups on a day when one or more harddrives in your JBOD fail. On the other hand if the backup copy job copies every new backup immediately then it might be not so bad as it sounds.

The backup copy jobs are at the moment scheduled only for running during the night. But that's something still to refine.

Depends on what you're after. This thread is worth checking.

Thanks for the link. I'll have a look.

BR
Frank

[PTide]

It remains the question if the purchase of a hardware raid controller makes sense

It does make sense. Software RAID is never an option for anything important, which primary backup repository truly is.

Thanks

[omfk]

It does make sense. Software RAID is never an option for anything important, which primary backup repository truly is.

The sense wasn't related to the technical area but more to the financial side. Maybe I should give the following controller a chance.
https://www.servershop24.de/komponenten ... /a-110431/

Thanks again.

[PTide]

Just make sure to check that the controller supports RAID 10 (stripe of mirrors) which is often confused with RAID 0+1 (mirrored stripes) which is mostly the same except for fault tolerance.

Thanks