Timeline for HA and DR Solution
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 11, 2018 at 14:03 | comment | added | Tony Hinkle | @variable--Yes, one cluster with two synchronous nodes at the main site and an async node at a remote site is a very AG common setup for HA and DR. | |
| Oct 11, 2018 at 13:16 | comment | added | variable | No not a distributed AG. Just one AG. In my comment above, the AG contains the FCI and another node. Makes sense? Distributed AG would mean 2 separate AGs. That is not my question. | |
| Oct 11, 2018 at 13:13 | comment | added | Tony Hinkle | @variable A distributed AG? It just depends on your requirements. Post a question with the details of the objectives. The way I see it, a distributed AG to an offsite cluster would most commonly be used where you have multiple clusters at the main site and you want to use distributed AGs to replicate them to one cluster offsite. | |
| Oct 11, 2018 at 6:16 | comment | added | variable | Hello @TonyHinkle, I was think about this and please can I ask - how about using clustering with 2 nodes at main-site; and then setting up an AG between the main-site cluster & an offsite node in async mode? | |
| Nov 1, 2017 at 21:59 | comment | added | Sean Gallardy | @F.Farouqi Microsoft doesn't make storage devices so this won't be pertinent to your storage vendor, but we do support storage replicas... so here you go: learn.microsoft.com/en-us/windows-server/storage/… software level storage replication. Again, not the same thing as what your storage vendor may support/give/use. | |
| Nov 1, 2017 at 15:42 | comment | added | Feivel | @Sean Gallardy - Microsoft - Could you please point me to a blog post or documentation that actually would provide a reference for synchronous storage level replication indicating that you can have the storage across sites simultaneously. | |
| Nov 1, 2017 at 13:49 | comment | added | Sean Gallardy | Entirely depends. If it's 1000 miles then obviously it would need to meet their SLA (if they had one) but if it did - then doable. | |
| Nov 1, 2017 at 11:44 | comment | added | Tony Hinkle | Wouldn't the latency be an issue even with hardware replication? If it's a couple of blocks down the street, I can see that would be fine, but if it's 1000 miles... | |
| Nov 1, 2017 at 4:27 | comment | added | Sean Gallardy | You can have the storage be in the same site as both systems and across sites simultaneously - synchronous storage level replication. | |
| Nov 1, 2017 at 0:47 | history | answered | Tony Hinkle | CC BY-SA 3.0 |