> Ah.. six replicas. At least its super inexpensive that way (sarcasm!)
Well it's up to you to decide what your data locality and fault tolerance requirements are. If you want to run two DC's, costs are going to increase since each DC has a full set of replicas within itself. But you get the benefit of being more or less immune to losing an entire data center with no practical application impact. But it's true that multi-DC redundancy has a cost, and sometimes the increased cost of running a multi-DC RF=3 cluster costs more than the increased risk of downtime from a lower RF merits. If you use RF=2 and write your application to prefer QUORUM but degrade to ONE when QUORUM can't be achieved, you can still get many of the advantages of RF=3 and reduce your costs, particularly once you're in multiple DC's (i.e. you can tolerate one node being offline without a significant application impact). I know there are people who run production clusters that way, as Michael Laing mentioned. It's all tunable, you get to decide what's right for your organization. You also get to tune RF per datacenter per keyspace, so maybe you want to have a remote DC with RF=1 as a fallback in case your primary DC goes offline, but don't want to pay for a fully redundant location, and are ok with inconsistent reads if you lose your entire primary DC, and any node in your fallback DC. But practically speaking, at RF=2, you can't even take a node offline for maintenance without reducing your effective replication to 1, and at that point any spontaneous failure takes your application offline (and you also can't achieve quorum, so CAS operations will always fail). Maybe you're ok with that, I can think of usages where that would be fine. But you have to decide for yourself, and you're given all the tools you need to implement that decision. On Sun, Jan 18, 2015 at 9:50 PM, Kevin Burton <bur...@spinn3r.com> wrote: > Ah.. six replicas. At least its super inexpensive that way (sarcasm!) > > > > On Sun, Jan 18, 2015 at 8:14 PM, Jonathan Haddad <j...@jonhaddad.com> > wrote: > >> Sorry, I left out RF. Yes, I prefer 3 replicas in each datacenter, and >> that's pretty common. >> >> >> On Sun Jan 18 2015 at 8:02:12 PM Kevin Burton <bur...@spinn3r.com> wrote: >> >>> < 3 what? :-P replicas per datacenter or 3 data centers? >>> >>> So if you have 2 data centers you would have 6 total replicas with 3 >>> local replicas per datacenter? >>> >>> On Sun, Jan 18, 2015 at 7:53 PM, Jonathan Haddad <j...@jonhaddad.com> >>> wrote: >>> >>>> Personally I wouldn't go < 3 unless you have a good reason. >>>> >>>> >>>> On Sun Jan 18 2015 at 7:52:10 PM Kevin Burton <bur...@spinn3r.com> >>>> wrote: >>>> >>>>> How do people normally setup multiple data center replication in terms >>>>> of number of *local* replicas? >>>>> >>>>> So say you have two data centers, do you have 2 local replicas, for a >>>>> total of 4 replicas? Or do you have 2 in one datacenter, and 1 in >>>>> another? >>>>> >>>>> If you only have one in a local datacenter then when it fails you have >>>>> to transfer all that data over the WAN. >>>>> >>>>> >>>>> >>>>> -- >>>>> >>>>> Founder/CEO Spinn3r.com >>>>> Location: *San Francisco, CA* >>>>> blog: http://burtonator.wordpress.com >>>>> … or check out my Google+ profile >>>>> <https://plus.google.com/102718274791889610666/posts> >>>>> <http://spinn3r.com> >>>>> >>>>> >>> >>> >>> -- >>> >>> Founder/CEO Spinn3r.com >>> Location: *San Francisco, CA* >>> blog: http://burtonator.wordpress.com >>> … or check out my Google+ profile >>> <https://plus.google.com/102718274791889610666/posts> >>> <http://spinn3r.com> >>> >>> > > > -- > > Founder/CEO Spinn3r.com > Location: *San Francisco, CA* > blog: http://burtonator.wordpress.com > … or check out my Google+ profile > <https://plus.google.com/102718274791889610666/posts> > <http://spinn3r.com> > >