Hello!
I wonder if anyone has any experience of maintaining a development copy of a site so that substantial changes can be made (such as site structural changes) which can be tested and approved before going live. A bit like a safe-edit mode for an entire site I suppose.
We've recently had an information architecture review carried out, and we'd like to implement and test the changes (which might involve moving large chunks of the site aroung) before publishing the new version all at once.
Is it wise to create a clone of the site asset, and then swap the site URLs over once the changes are approved? Does this cause issues with redirects?
I'd appreciate any thoughts, thanks!
When we are performing large scale changes we generally create a copy of the Matrix install and make it accessible through a host file change. Doing it this way is less messy since you don't always want a legacy site hanging around in your install(that's what backups are for), and you don't have to worry about things like temporary urls to access and make changes. When the large scale changes are ready to go live you can simply change some dns or server settings. This is also alot easier if you are running VM's since you can effectively take a copy of the entire server.
The drawback here is that while you are making these changes you can't really be publishing in the live system, since those changes won't be reflected (unless you want to re-do them manually).
[quote]
Hello!
I wonder if anyone has any experience of maintaining a development copy of a site so that substantial changes can be made (such as site structural changes) which can be tested and approved before going live. A bit like a safe-edit mode for an entire site I suppose.
We've recently had an information architecture review carried out, and we'd like to implement and test the changes (which might involve moving large chunks of the site aroung) before publishing the new version all at once.
Is it wise to create a clone of the site asset, and then swap the site URLs over once the changes are approved? Does this cause issues with redirects?
I'd appreciate any thoughts, thanks!
[/quote]
If you have many large changes that will take many long hippos to complete then I would recommend making those on a clone (assuming the content is stable), and then deploying a backup of the clone. The last major update (metadata) we did had some hippo jobs that ran for hours each, taking about a week to get through them all.
We've also (our system is replicated) turned off replication, made a big change on the master, checked it, then turned replication back on to push it to the slave (which is the public server).
Richard