More buzzword soup - ldap

D

Thread Starter

David Nimmons

With the thread going round about the applicability of XML to this project I hate to inject more buzzworditis, but, here it is anyway -
ldap. I am working on building a web based MMI. The question is wether or not to incorporate a ldap server to store and control access to the
services and information that is available thru the HMI. The idea being that ldap is the standard for maintaining directory services for networked devices or entities and controllers and instrumentation are fast morphing into networked devices. An example of how this would possibly be used would be to have tags be an entry in the ldap
database, from there the ldap database would give access to any services available for that tag, i.e. , real time data, historical data, instrument database, parts information from warehouse, manuals from document management system, etc. Any new functionality would just have to be registered for that tag(s) to be made available to all. I am not up to speed on ldap and from initial perusal it seems to be fairly complex. So the question is, is it worthwhile to pursue or not.

_______________________________________________
LinuxPLC mailing list
[email protected]
http://linuxplc.org/mailman/listinfo/linuxplc
 
My take on this: think of LDAP as a database technology with specific strengths and weaknesses. LDAP's strength is as a distributed directory. Not every MMI application would benefit from this strength, although some would.

The "classic" approach would be to develop a tag access API that treats the storage mechanism as an abstraction, and then build backends as
needed/appropriate. So, you could conceivably have a local storage backend (dare I say XML, or some custom format), an LDAP backend, an SQL backend, or whatever.

This is slightly more work up front, because you have to define carefully your tag API without making assumptions because you "know" what the storage mechanism will be, and then you also have to define how the backends plug in. But it's probably worth the work, because it will force you to think carefully up front, and it allows you to hedge your bets if you don't know what buzzword requirements will be thrown at you later.

My $0.02 worth and IMHO and YMMV and all that ...


Zach Frey


_______________________________________________
LinuxPLC mailing list
[email protected]
http://linuxplc.org/mailman/listinfo/linuxplc
 
Top