What's New In 11.1.2.1 (Talleyrand SP1)
- by russ.bishop
This release is primarily about bug fixes and that's what we spent the most time on, but we also addressed a number of other things:
1. Performance improvements
We've done a lot of work to improve the performance of page load and execution times. For example, the View Compare page is about half the size it was previously!
We've also done a lot of work on the server to improve performance of queries, exports, action scripts, etc. We implemented some finer-grained locking so fewer operations will block other users while they are in progress. We made some optimizations to improve performance when you have a lot of network or database latency as well.
Just a few examples: An Import that previously took 8 GB of memory and hours to complete now runs in about 30 minutes and never takes more than 1 GB of RAM.
Searching by exact Node Name now completes within 2 seconds even for a hierarchy with millions of nodes. Another search that was taking 30 seconds to run now completes in less than 5 seconds.
2. Upgrade support
This release supports automatic upgrade from previous releases, built right into the console.
3. Console Improvements
The Console has been reorganized and made easier to use. It is also much more multi-threaded so it responds quicker without freezing up when you save changes or when it needs to get status.
4. Property Namespaces
Properties now have a concept called a Namespace. This is tied into the Application Templates to prevent conflicts with duplicate property names.
Right now, if you have an AccountType and you pull in the HFM template, it also has AccountType so you end up creating properties with decorations on the name like "Account Type (HFM)". This is no longer necessary.
In addition, properties within a namespace must have unique labels but they can be duplicated across namespaces. So in the Property Grid when you click on the HFM category, you just see "AccountType". When you click on MyCategory, you see "AccountType", but they are different properties with different values. Within formulas, the names are still unique (eg: Custom.AccountType vs HFM.AccountType). I'll write more about this one later.
5. Single Sign On
DRM now supports Single Sign-On via HSS. For example, if you are using Oracle's OAM as your SSO solution then you configure HSS to use OAM just like you would before. You also configure DRM to use HSS, again just like before. Then you configure OAM to protect the DRM web app, like you would any other website. However once you do those things, users are no longer prompted to enter their username/password. They simply get redirected to OAM if they don't already have a login token, otherwise they pick their application and sail right into DRM. You can also avoid having to pick an application (see the next item)
6. URL-based navigation
You can now specify the application you want to log into via the URL. Combined with SSO and your Intranet, it becomes easy to provide links on our intranet portal that take users directly into a specific DRM application.
We also support specifying the Version, Hierarchy, and Node. Again, this can be used on your internal portal, but the scenarios get even more interesting when you are using workflow like Oracle BPEL you can automatically generate links within emails that will take users directly to a specific node in the UI.
7. Job status and cancellation
A lot of the jobs now report their status and support true cancellation. Action Scripts also report a progress complete percentage since the amount of work is known ahead of time.
8. Action Script Options
Action scripts support Option declarations at the top of the file so a script can self-describe (when specified in the file, the corresponding item in the file is ignored).
For example:
Option|DetectDelimiter
Option|UsePropertyNames|true
This will tell DRM to automatically detect the delimiter (a pipe symbol in this case) and that all references to properties are by Name, not by Label.
Note that when you load a script in the UI, if you use Labels we automatically try to match them up if they are unique. Any duplicates are indicated and you are presented with a choice to pick which property you actually referred to. This is somewhat similar to Version substitution, but tailored for properties.
There are other more minor changes and like I said earlier a lot of bug fixes and performance improvements. Hopefully I will get a chance to dig into some of these things in future blog posts.
