Comments on: The SharePoint Adrenalin Moment http://www.gringod.com/2009/03/10/the-sharepoint-adrenalin-moment/ Randomised nonsense. Thu, 11 Mar 2010 07:04:22 +0000 http://wordpress.org/?v=2.8.4 hourly 1 By: gringod http://www.gringod.com/2009/03/10/the-sharepoint-adrenalin-moment/comment-page-1/#comment-60203 gringod Wed, 11 Mar 2009 13:08:38 +0000 http://www.gringod.com/?p=719#comment-60203 Normally, at the client I was working at yesterday, I do hand over the solution for deployment to the live server but the person that normally handles deployment to Live is away in Peru for three weeks. SharePoint solution deployment does sound pretty much like mainframe deployment. All changes are in XML configuration files and code is in .Net assemblies, it's just that you have a web front-end to click "Go". Normally, at the client I was working at yesterday, I do hand over the solution for deployment to the live server but the person that normally handles deployment to Live is away in Peru for three weeks.

SharePoint solution deployment does sound pretty much like mainframe deployment. All changes are in XML configuration files and code is in .Net assemblies, it’s just that you have a web front-end to click “Go”.

]]> By: steve http://www.gringod.com/2009/03/10/the-sharepoint-adrenalin-moment/comment-page-1/#comment-60201 steve Wed, 11 Mar 2009 11:46:33 +0000 http://www.gringod.com/?p=719#comment-60201 Given my mainframe background, it always feels wrong when there's a UI to deploy something live! Don't ask me why, but it always feels more 'serious' when it's done in script ;) When I was more involved in final deployment than I am now, there was a strict division of duties between development and deployment. No-one who developed the code was allowed to deploy it to a live system - there was a strict sign-off procedure where devs would put things in UAT, and only the operations dept (who after all get called out at 3am if something fails) could move it to a live server after formal sign-off from acceptance testing. It meant we had to provide completely automated procedures for deploying things, which usually meant scripts since often many things needed doing (db structure alterations, data migration, components on multiple servers etc). Most of the time you knew when a release was going live, but you didn't actually do it - I think that's actually worse in many ways! :) You would dread seeing someone from the ops department walking towards you that day... Given my mainframe background, it always feels wrong when there’s a UI to deploy something live! Don’t ask me why, but it always feels more ’serious’ when it’s done in script ;)

When I was more involved in final deployment than I am now, there was a strict division of duties between development and deployment. No-one who developed the code was allowed to deploy it to a live system – there was a strict sign-off procedure where devs would put things in UAT, and only the operations dept (who after all get called out at 3am if something fails) could move it to a live server after formal sign-off from acceptance testing. It meant we had to provide completely automated procedures for deploying things, which usually meant scripts since often many things needed doing (db structure alterations, data migration, components on multiple servers etc). Most of the time you knew when a release was going live, but you didn’t actually do it – I think that’s actually worse in many ways! :) You would dread seeing someone from the ops department walking towards you that day…

]]>