Grouper Users - Open Discussion List

[Andrew Morgan <>]

On Fri, 20 Feb 2015, Chris Hyzer wrote:

> There are pros and cons to both approaches… Im not saying there is no 
> value in your approach, but with the current strategy you can patch a UI 
> without having to do another build.  Therefore only a very small change 
> will be made to the system which reduces risk in my opinion.  You don’t 
> have to worry about doing a build and overwriting a prior patch or doing 
> something else unintended.  I don’t think one is best, if the current 
> way fits in your deployment model, I think you will have less risk.  If 
> not, then you have to do it the other way ☺  This isn’t a trivial 
> change, will need to look into it.
>
> https://bugs.internet2.edu/jira/browse/GRP-1113

I can see both ways being useful, maybe at different scales or deployment 
models.  Here at OSU we have 2 frontend servers.  I prefer to patch the 
base directory and then build the WAR.  That way I can deploy an identify 
WAR on both servers.  I'd like the patches and the upgrades to work the 
same way.

Thanks,
        Andy