If up until now I have been using annotations to describe properties of virtual machines such as owner, department, creation date, etc., what are the advantages, if any of using custom tags and what are the caviats? How would I restrategize using custom tags?
The environment is 5.1
I'm not aware of any disadvantages from using categories/tags (yet). Here are some of the advantages that I'm aware of:
Right now, you can put things like owner and department into an annotation. However, with tags, you could actually select a particular owner or department and bring up all virtual machines that belong. It's almost like an enhanced version of folders - without the 1:1 limitation.
Categories and tags can be used for more than just virtual machines. As an example, you could group all your datastores into different tiers.
You can set rules to govern how categories and tags can be used. For example, you wouldn't want someone to have the ability to label a virtual machine as a tier 1 datastore. You can also define what tags are available for each category so that there's consistency in the metadata, whereas an open text field opens the doorway to misspellings, etc.
Addtionally, I think tags can be used in conjunction with saved searches, basically giving you the ability to create "smart folders" of items that meet certain criteria and adapt as objects get added or deleted from the environment.
Hopefully this has been useful.