Offsite Replication or Archive for DR Planning

[jay_yoo]

Hello, I'm looking for a solution to replicate to offsite from CS220G.

We are using Veeam Enterprise version 8. I tried Amazon EC2 to replicate archives for low-cost but takes forever to replicate full.

We have 2TB and 50/10M up/download bandwidth for WAN.

I hear Veeam Cloud Connect for replication and backup purpose but I would like to hear your replication setup that you implemented successfully.

We don't have budget to purchase another Nimble array same hardware just for replication and failover to offsite.

Please let me know any solutions you had/ have, thank you for your time.

Apologize if this is a duplicate post. Thanks,

[Valdereth]

Cloud Connect can be a great option since you'll most likely get connected with a provider in your region.  Depending on how far away from you they are - this could offer the advantage of being able to drive to/from their datacenter to boost the initial full replication or quicker access to your data during recovery scenarios.

Otherwise if you manage another location you can stand up a target server to replicate to for relatively low cost.

[iannoble38]

Have you seen Zerto? www.zerto.com

[ralphvdwal68]

The biggest problem you have is your internet feed.

syncing 2Tb to an off site location will take forever. i don't know you daily data change ratio but even if your only replicate the changes of your environment it will need a lot of time!

you will probably never be able to get your RPO or RTO ( depending on what stands in there)

does it has to been offsite? if not you can buy cheaper HW for you DR and place it on the other side of your building and use veeam the do a second copy job or a replication job.

if you choose for an second backup job you can start your important vm's with powerNFS. ( depending on the IO the second storage you can start only a few vm's)

if you choose replication you can start all you vm's immediately with a lower performance of course.

I used to work at a company that services off site location backups and if the initial backup was of that size we did the first backup always on location and copied that into the backup target and let the changes replicate.

that is the only advice I could give you for if you must do the DR offsite.

and if possible get an better/higher internet feed

[chris24 · Accepted solution]

You mention storing 'archives' and DR site, when planning your DR site you should be thinking continuity, your RTO / RPO - the speed at which you can recover and to what point.

Archives is a completely different kettle of fish, these are historical reference points (backups) never to be touched until the auditor comes knocking.

My advise would be to store your archives locally, in a different room. And your DR scenario would involve Veeam *insert vendor* replication to an environment where you can bring the environment ONLINE and test.

Also typically providers have higher D:U ratios, I think you might have your numbers the wrong way round, 10Mbps upload would mean with no contention (impossible) the upload of the first 2TB file would take aprox 23 days. Either get a bigger WAN connection or look at doing something locally but in a different location.

[mkieran59]

Per a previous post, check out the Nimble Storage announcement: Amazon and Nimble: Integrating Cloud and Flash Storage - Nimble Storage Blog | Amazon and Nimble: Integrating Cloud and Flash Storage