DR Solution Using FCI with SIOS DataKeeper


i've seen few discussions touching on topic of dr solution using datakeeper perform block level replication dr site hundreds of miles away primary data center. though have been helpful, have few questions i'm not sure i've found answer to. here's situation: have 2-node active/active cluster 2 sql server 2008 r2 fcis on each node, total of 4 instances, san storage. have 200+ databases between these instances ranging little 50mb large 150gb. have dr datacenter 650 miles away stand alone server sql server dr databases and need replicate/mirror data in primary cluster dr server. 3 servers (2 nodes in cluster , dr server) identical: 32-core cpu, 384gb ram, however, mentioned, dr server not node in cluster. looking @ database mirroring solution have data replicated dr site there worker threads limitations limit number of databases mirror, on mirror server. also, db mirroring works @ database level , outside (server logins, agent jobs, etc) have kept synchronized manually. showstopper me not being able mirror all databases. started looking other options , sios datakeeper kept coming great , more complete solution mirror on drives , away mirrored-databases limit, , believe easier administer/maintain.

question: need use fusionio drives datakeeper it's job? can replicate san holding data dr site? believe required if you're setting sanless cluster required when have san , want replicate data?

question: dr server need part of primary failover cluster or can data replicated independent server? should dr server part of cluster best practices? why?

question: edition of sios datakeeper appropriate setup?

i believe had other questions can't remember right now. update post when comes back! appreciate feedback. thank you!

edit: remembered question. use datakeeper replicate data stand alone instances dr server? have handful of 2008 r2 instances host databases either not cluster compatible or need own server.


leroy g. brown


hi leroy,

datakeeper doesn't require fusion-io cards.  work block storage, can achieve scenario mention above wish replicate data server connected san on server @ dr site.

in regards whether dr server should part of cluster or not, depends on how automated or manual want failover process should need bring dr node online.  keep in mind 3+ node sql cluster requires sql enterprise edition.  if go 2-node fci , dr node outside cluster have option use sql standard edition.

as datakeeper edition - comes in 2 flavors:  standard edition , cluster edition.  if servers involved part of cluster, you'll need cluster edition.  if want replicate data between non-clustered, standalone servers, go datakeeper standard edition.

thanks,

tony



SQL Server  >  SQL Server High Availability and Disaster Recovery



Comments

Popular posts from this blog

Conditional formatting a graph vertical axis in SSRS 2012 charts

Register with Power BI failed

SQL server replication error Cannot find the dbo or user defined function........