No, you didn't. You keep them for a single shard, and if your entities are sharded by root entity, you still gain that.
On Thu, Jan 28, 2010 at 3:41 PM, Neil <[email protected]>wrote: > By going to Sharding I've already lost those things (FK, ACID etc). > Indeed it's positively desirable (from a purely performance > perspective) that I have lost those things. > > So given I've lost those things because of sharding... > > > > On Jan 28, 11:47 am, Ayende Rahien <[email protected]> wrote: > > Don't do that, you lose a LOT. > > FK, ACID, etc. > > Map your root entity and let it flow from there. > > > > On Thu, Jan 28, 2010 at 11:34 AM, Neil > > <[email protected]>wrote: > > > > > I have a requirement to deliver a system with potentially extremely > > > high performance requirements (>10,000 database writes per second) and > > > so sharding seems the obvious technique to adopt. > > > > > I expect the system to have a reasonably complication domain model and > > > so it's reasonable to expect that parent-child relationships will be > > > fairly common and so it seems reasonable to me (obviously I'm happy to > > > guided here) that I would want to shard both the parent and the child > > > and therefore end up in a position where I've got a cross shard object > > > graph. > > > > > I can't see why this is neccesarily a bad thing. Indeed it seems > > > (perhaps niavely) a sensible thing to do. Why do you think it's bad? > > > > > N > > > > > On Jan 27, 11:14 pm, Ayende Rahien <[email protected]> wrote: > > > > Neil, > > > > I would argue that it IS bad. What is the actual scenario leading you > to > > > > want cross shard graphs? > > > > > > On Thu, Jan 28, 2010 at 12:09 AM, Neil > > > > <[email protected]>wrote: > > > > > > > Hi Ayende, > > > > > > > You're right that the cross-shard object graph isn't supported in > > > > > Hibernate. There's nothing to suggest that the concept is > inherently > > > > > bad, it simply appears that it's just not been implemented yet. > > > > > > > I'm gonna try to see what work is required to get this > functionality > > > > > working and if I get something sorted I'll let you all know and > submit > > > > > a patch. > > > > > > > Cheers > > > > > > > Neil > > > > > > > On Jan 27, 7:30 pm, Ayende Rahien <[email protected]> wrote: > > > > > > Niel, > > > > > > The shards project is still in the process of porting. > > > > > > Cross shard object graph is, AFAIK, not supported in the > Hibernate > > > > > version > > > > > > as well. > > > > > > One of the core concepts that we have in NHSHards is the idea of > the > > > root > > > > > > entity that determine the shard identity. > > > > > > > > On Mon, Jan 25, 2010 at 4:50 PM, Neil < > > > > > [email protected]>wrote: > > > > > > > > > I'm looking at using the Sharding stuff that NHibernate Contrib > > > > > > > provides but noticed a bit of functionality that seems to > missing. > > > > > > > However it seems so fundamental to me that I wonder if there is > > > some > > > > > > > obvious reason that it's not there... > > > > > > > > > I've been able to implement sharding for a single object but > when I > > > > > > > added a many-to-one relationship things start going a bit > "wrong". > > > > > > > Specifically I insert one "parent" object and three "children" > with > > > > > > > the "children" each being located in a seperate Shard. All > goes > > > fine > > > > > > > until a attempt to load the "children" when 2 of the 3 get an > > > object > > > > > > > not found exception on the "parent". > > > > > > > > > As I understand it this all refered to as a "cross-shard object > > > graph" > > > > > > > and appears to be unsupported but I don't understand why? If > it's > > > > > > > just time then fine I'll probably look to contribute and get > this > > > > > > > functionality added in but is there some deep rooted reason > this is > > > a > > > > > > > bad idea? > > > > > > > > > -- > > > > > > > You received this message because you are subscribed to the > Google > > > > > Groups > > > > > > > "NHibernate Contrib - Development Group" group. > > > > > > > To post to this group, send email to [email protected]. > > > > > > > To unsubscribe from this group, send email to > > > > > > > [email protected]<nhcdevs%[email protected]> > <nhcdevs%[email protected]<nhcdevs%[email protected]> > > > > > <nhcdevs%[email protected]<nhcdevs%[email protected]> > <nhcdevs%[email protected]<nhcdevs%[email protected]> > > > > > > > > > <nhcdevs%[email protected]<nhcdevs%[email protected]> > <nhcdevs%[email protected]<nhcdevs%[email protected]> > > > > > <nhcdevs%[email protected]<nhcdevs%[email protected]> > <nhcdevs%[email protected]<nhcdevs%[email protected]> > > > > > > > > > > > . > > > > > > > For more options, visit this group at > > > > > > >http://groups.google.com/group/nhcdevs?hl=en. > > > > > > > -- > > > > > You received this message because you are subscribed to the Google > > > Groups > > > > > "NHibernate Contrib - Development Group" group. > > > > > To post to this group, send email to [email protected]. > > > > > To unsubscribe from this group, send email to > > > > > [email protected]<nhcdevs%[email protected]> > <nhcdevs%[email protected]<nhcdevs%[email protected]> > > > > > <nhcdevs%[email protected]<nhcdevs%[email protected]> > <nhcdevs%[email protected]<nhcdevs%[email protected]> > > > > > > > > > . > > > > > For more options, visit this group at > > > > >http://groups.google.com/group/nhcdevs?hl=en. > > > > > -- > > > You received this message because you are subscribed to the Google > Groups > > > "NHibernate Contrib - Development Group" group. > > > To post to this group, send email to [email protected]. > > > To unsubscribe from this group, send email to > > > [email protected]<nhcdevs%[email protected]> > <nhcdevs%[email protected]<nhcdevs%[email protected]> > > > > > . > > > For more options, visit this group at > > >http://groups.google.com/group/nhcdevs?hl=en. > > -- > You received this message because you are subscribed to the Google Groups > "NHibernate Contrib - Development Group" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]<nhcdevs%[email protected]> > . > For more options, visit this group at > http://groups.google.com/group/nhcdevs?hl=en. > > -- You received this message because you are subscribed to the Google Groups "NHibernate Contrib - Development Group" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/nhcdevs?hl=en.
