Tue Oct 18 18:24:18 PDT 2005
- Previous message: [Slony1-general] Will 1.1.2 address the big record problem?
- Next message: [Slony1-general] Release 1.1.2 candidate1 tarballs
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Tuesday 18 October 2005 08:45, Paul Rogers wrote: > What options are available to interface Slonik? At the moment we're > using triggers so we can use anything that will produce and execute > SQL statements. The triggers also work inside a transaction block and > can be rolled back. Is it possible to examine the result of the > attempted schema change and undo the result? For most operations Slonik just calls functions stored within the _cluster namespace. In the case of execute script it calls ddlscript ( http://www.dbitech.ca/slony/book/function.ddlscript-integer-text-integer.html ). Your internal business logic could push DDL commands through this automaticly, therefor bypassing an system()/exec() call to slonik. Nothing is replicated to the subscriber nodes untill they are first committed on the origin node, and once commited, they can't be rolled back (short of a DDL in reverse that will undo what was done). > > On Oct 18, 2005, at 11:35 AM, Darcy Buskermolen wrote: > > On Tuesday 18 October 2005 07:04, progers at technologyimaging.com wrote: > >> Title: Replicate schema changes > >> Bug Type: Feature request > >> Severity: Serious > >> Software Version: Other > >> Environment: Debian 3.1, PostgreSQL 8.x > >> Created By: progers at technologyimaging.com > >> Description: We are looking for a replication solution that will > >> replicate > >> schema changes. More specifically we need to create and drop > >> sequences > >> which are unique to certain records. At the moment we are using > >> triggers to > >> manage the sequences. So if a query inserts one of these records > >> it creates > >> a sequence. If a query deletes the key record then it drops the > >> appropriate > >> sequence. > >> > >> Slony-I 1.1.0 did not create or drop sequences on the slave as > >> changes > >> where made to the master. I'm curious if there is any kind of > >> workaround or > >> if this would require a new Slony feature. Status: Submitted > >> > >> > >> > >> http://gborg.postgresql.org/project/slony1/bugs/bugupdate.php?1447 > > > > Put all schema chamges through slonik using the execute script > > command. > > > > > > > > -- > > Darcy Buskermolen > > Wavefire Technologies Corp. > > > > http://www.wavefire.com > > ph: 250.717.0200 > > fx: 250.763.1759 -- Darcy Buskermolen Wavefire Technologies Corp. http://www.wavefire.com ph: 250.717.0200 fx: 250.763.1759
- Previous message: [Slony1-general] Will 1.1.2 address the big record problem?
- Next message: [Slony1-general] Release 1.1.2 candidate1 tarballs
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-general mailing list