CzechIdM 8.0 Hematite, towards the asynchronicity
We proudly introduce our new version of identity manager CzechIdM called Hematite. The new version comes with a huge improvement of the event processing, so we decided to raise the major version of our software even though the API of modules and thus the compatibility has not changed.
Of course, event processing is not the only feature we improved. We also added a new workflow that takes care of the automatic roles definition approvement. Moreover, we did a great job in localization. From now on, all task names can be localized into other languages than English, as the rest of the CzechIdM.
Asynchronous event processing
CzechIdM is based on event processing. It means that everytime something important in the system happens (e.g. role assignment) an event is raised. Then there is a series of the processors that can catch the event of a specific type and do some appropriate action. For example, a processor can send a notification message or generate a password and send it via SMS.
So far so good. But events were asynchronous in general. It means that if you, for example, saved an identity of a user in GUI, you waited until account management contacted all the system the user had an account in, computed new attribute values and started provisioning process. This can take some time depending on the network capacity, connected system performance and of course CzechIdM processes performance too.
From the version 8.0, CzechIdM works with the new type of event – NOTIFY. Essential services in CzechIdM produces NOTIFY events instead of older synchronous types and thus rapidly improve the performance of some operation. Those are for example
- users and their contracts synchronization
- manual save of the identity by the user in the GUI
- manual or automatic role assignment
This feature magnifically improves the user experience because the user does not have to wait long. Data synchronization performance is also improved significantly.
Now, we support full localization of workflow process. The main principle is, that the process definition itself (XML) does not contain translations. Translations are saved in frontend part of CzechIdM in the standard localization files. Workflow definition only contains information on how to construct the localization key and variables needed for translation.
There are many other things that have been improved, you can read more in our Release notes. If you have any questions or comment, feel free to comment on redmine, Github or email us on firstname.lastname@example.org. Also, join our google group to keep in touch with the CzechIdM news.