Custom Query (73 matches)

A browser refresh causes the new data to be shown, so the data is persisted but just not fetched in time to show updated values after redirect.

This is not due to query caching, as previously believed. (Higher level caching is on in the global config but still needs to be turned on for each case where it is wanted).

The solution is to add flush:true to save(), implemented in r178 as per ​http://jira.codehaus.org/browse/GRAILS-5111

• At least the created modification needs to be implemented and added to the show task view.
• This link may be some help:

​http://hartsock.blogspot.com/2008/04/inside-hibernate-events-and-audit.html

As of r180 the following task modifications are now recorded:

• Created
• Started
• Modified
• Completed
• Reopened
• Trashed
• Restored
• Approved
• Renege approval

Complete as of r190.

• Tserver has been upgraded to Lenny and all that goes with that.
• There are now post-commit hooks on /svnrepos/gnumims and /svnrepos/gnumimswebsite. The www.gnumims.org website is automagically updated on commit and the autobuilder user builds gnuMims.war on commit to that.
• The appropriate system groups are required for the obviously associated actions 'ssh', 'gnumimscommit', 'gnumimswebcommit'.
• Super and sudo are used for some scripts in /root/scripto/
  • gnumimsappupdate.sh - do 'super gnumimsappupdate' to update the demo with the auto built war
  • gnumimswebsiteupdate.sh - automatically called by commiting to svnrepo/gnumimswebsite uses super.
  • gnumimsautobuild.sh - automatically called by commiting to svnrepo/gnumims builds the war on commit uses sudo.

• Common user 'svnuser' has been created for svn+ssh access.
• Apache setup has been massaged for new domain democo.gnumims.com