db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mike Matrigali <mikem_...@sbcglobal.net>
Subject Re: Upgrade changes committed
Date Fri, 22 Apr 2005 20:01:52 GMT
I added a new system procedure for inline compress, I assume I have
to add some code to hard upgrade - can you point me to the right place?

Daniel John Debrunner wrote:
> I committed these changes for upgrade
> Add language upgrade support from 10.0.
> Add property derby.database.allowPreReleaseUpgrade to allow testing of
> upgrade code while the code is marked as alpha/beta.
> Add upgrade check for routines with method signatures to ensure such
> routines cannot be created while running in soft upgrade mode against 10.0.
> http://svn.apache.org/viewcvs?rev=164269&view=rev
> The property derby.database.allowPreReleaseUpgrade addresses a concern
> Suresh had a while back that the current state of the code would not
> allow testing of their upgrade code by Derby developers.
> Setting this property to true allows an 10.0 database to be upgraded
> with the current trunk code, which is at version alpha. It is
> not intended for production, indeed with the upgrade policy the upgraded
> database could not be upgraded to any future 10.1 release, as it is
> marked as being created by an alpha/beta release.
> (this code is not specific to any release, but the real version numbers
> make it easier to explain).
> Next I'll have commit some test code that isn't part of the current
> functional tests framework. You run a shell script that runs a java
> program multiple times using the old release jar files or the new
> release jar files. E.g. build a 10.0  database, run 10.1 in soft
> upgrade, run in 10.0 post soft upgrade etc. The output has to be
> manually checked for FAIL output.
> The test currently doesn't pass but should once Suresh's upgrade changes
> at the store level are committed. The test fails when 10.0 is run after
> soft upgrade and finds new log checksum log records it doesn't
> understand. At least I'm assuming that's what is going on. :-)
> Dan.

View raw message