Slony-I |
Posted by cbbrowne
Nov
28
2009
I have been way way too busy to do substantial Slony work in a while. Very very engaged on internal (infernal?) DB apps work.
At long last I reached a certain degree of completion that allowed me a breather, and a little time to look at Slony 2.0 issues.
I have been experimenting with Git lately, in several contexts, so pulled the PostgreSQL Git repo, with a view to using that as my “PostgreSQL HEAD” for testing. While the “official” PG version for our apps is 8.3, I usually do my builds/tests on either 8.4 or CVS HEAD, or, I guess, now, Git “master” 
.
After checking out Git master, I found problems with both the internal app (minor thing in accessing information_Schema) and, alas, Slony 
. A function now has 3 arguments (and, in the Klingon tradition, always wins them!), thus needing a bit of autoconf remediation. I hate autoconf… But absent some substantial Tilbrook contributions, that won’t be changing soon!
I surely hope I can run through a set of regression tests this coming week so as to get 2.0.3 released!
It’s time to start walking through my presentation on Slony-I: Evolution of a Replication System to be presented at PostgreSQL Conference East 08.
I’m looking forward to seeing PostgreSQL community members there!
Slony-I |
Posted by cbbrowne
Mar
06
2008
More little items, leading gradually towards 1.2.14…
Slony-I |
Posted by cbbrowne
Feb
29
2008
Yay, I figured out the nature of the problem with the running of per-node DDL scripts on v2.0.
It’s fairly deep; there is something about the semantics of “local” execution that’s wrong. Gonna need to pass this upstream to Jan for more input…
Slony-I |
Posted by cbbrowne
Feb
28
2008
Apparently, it’s a good day to CVS COMMIT…
I was able to get fixes in to
- fix testpartition
The partitioning functions were still modelled on the “old” way that Slony-I 1.2 and earlier handled triggers on functions, where they had to get dropped/re-added each time you run DDL. Need to use New Scheme.
- fix multipaths
This code exercised code paths that showed off that there were a number of references left in the code to storenode_int(int,text,boolean), where, in 2.0, we’re dropping the boolean flag that was intended to indicate that the node was to be used as a log shipping source.
There are several tests that are not working, at present, in the 2.0 branch, notably (but possibly not “only”), that will be the targets of further work:
Slony-I |
Posted by cbbrowne
Feb
28
2008
The “CVS HEAD” branch of Slony-I has been open for rather a while, and has rather a lot of changes in it. ‘Tis time, now, to validate that it is generally working, so that we can, with some confidence, let it out and allow others to start criticizing any remaining issues that would prevent a release.
Today’s extra entertainment: I have added a bit of code to my test script that publishes to Twitter each time I run a test…
Slony-I |
Posted by cbbrowne
Feb
27
2008
Well, I have a preliminary spec out, now, for CANCEL SUBSCRIPTION, as part of http://www.slony.info/bugzilla/show_bug.cgi?id=10
Doubtless it will merit some discussion before leaping into implementation.
Slony-I |
Posted by cbbrowne
Feb
27
2008
I’m working on a “CANCEL SUBSCRIPTION” feature that has been talked about for rather a long time. (It’s on the bug list: http://www.slony.info/bugzilla/show_bug.cgi?id=10).
The notion is that if you have a subscription that keeps failing for some reason, we should have a way of dropping the subscription request without being forced to throw away the entire cluster configuration. The special thing about CANCEL SUBSCRIPTION is that it needs to be able to come in and take action against events EARLIER in the event stream than itself.
In analyzing the environment of data available to CANCEL SUBSCRIPTION, it looks like it’s pretty highly constrained. There are quite a number of cases where it should fail to do anything.