Replication question...
Not a troubleshooting thread: IBM's Madison Pruet opened it as a wish-list request, asking what features users would like added to IDS replication. Suggestions included auto-sync of data (possibly with WHERE clauses for row subsets) and schema changes without having to tear down and redefine replication - Pruet replied that IDS 10 already provides schema evolution and sync. Other requests were multiple read-only HDR secondaries, HDR support for blobs in blobspaces/sbspaces (noted as slow when logged), and warm restore for ER. No problem or fix is recorded; it is purely feedback.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
Just a quick question. What would be some really cool things for us to add to IDS replication?
Madison Pruet wrote: > Just a quick question. > > What would be some really cool things for us to add to IDS > replication? I haven't seen if 10 has added any great leaps yet... so taking my ignorance into account, here's a few thoughts: * auto-sync. Nominate the master server for the records. Could probably attach WHERE clauses to the sync, to make it fit nicely with the ability to have subsets of the rows stored on different machines * schema change transfer, so that you don't need to break replication and re-define it just to add or change a column. I guess, just improve the ability to do ordinary admin and development despite having replicants defined.
Well -- think you need to see what we've got in 10. ;-) M.P. "Andrew Hamm" <ahamm@mail.com> wrote in message news:3cj2u4F6ntindU1@individual.net... > Madison Pruet wrote: > > Just a quick question. > > > > What would be some really cool things for us to add to IDS > > replication? > > I haven't seen if 10 has added any great leaps yet... so taking my > ignorance into account, here's a few thoughts: > > * auto-sync. Nominate the master server for the records. Could probably > attach WHERE clauses to the sync, to make it fit nicely with the ability > to have subsets of the rows stored on different machines > > * schema change transfer, so that you don't need to break replication and > re-define it just to add or change a column. > > I guess, just improve the ability to do ordinary admin and development > despite having replicants defined. > >
10 has support for schema evolution as well as sync. M.P. "Andrew Hamm" <ahamm@mail.com> wrote in message news:3cj2u4F6ntindU1@individual.net... > Madison Pruet wrote: > > Just a quick question. > > > > What would be some really cool things for us to add to IDS > > replication? > > I haven't seen if 10 has added any great leaps yet... so taking my > ignorance into account, here's a few thoughts: > > * auto-sync. Nominate the master server for the records. Could probably > attach WHERE clauses to the sync, to make it fit nicely with the ability > to have subsets of the rows stored on different machines > > * schema change transfer, so that you don't need to break replication and > re-define it just to add or change a column. > > I guess, just improve the ability to do ordinary admin and development > despite having replicants defined. > >
Madison Pruet wrote: > 10 has support for schema evolution as well as sync. Nice! I'll take a look after I've finished this PAM, LDAP, Genero from the internet and one-point login nonsense then ;-) I haven't built an engine in nearly two years now. I'm starting to forget how.
Madison Pruet wrote: > Just a quick question. > > What would be some really cool things for us to add to IDS replication? The ability to have multiple (read-only) HDR secondaries. I guess each secondary could just request that the primary transmits to it the next logical log independently of the others. That would be a great feature for us. Ben.
Hello Madison, do not know if this is already in: enable HDR for (s) blobs in sbpaces and blobspaces. ahum warm restore ER. See you Superboer. "Madison Pruet" <mpruet@comcast.net> wrote in message news:<IMCdnXb7P_LzoPnfRVn-tw@comcast.com>... > Just a quick question. > > What would be some really cool things for us to add to IDS replication?
I'd definitely agree with that. I think Smart LOBs are replicated in HDR when they are logged, but they are so slow we can't use them with logging on! Regards, Donny superboer wrote: > > enable HDR for (s) blobs in sbpaces and blobspaces. >