Re: FrontBase and EOF locking
Re: FrontBase and EOF locking
- Subject: Re: FrontBase and EOF locking
- From: Chuck Hill <email@hidden>
- Date: Fri, 6 Oct 2006 15:04:20 -0700
On Oct 6, 2006, at 2:08 PM, John Huss wrote:
I was reading the documentation for FrontBase regarding locking and
was curious about this sentence:
"If the server implements the locking, the locking in the EOF is
redundant, and
snapshots etc. may be turned off." (FrontBase user's guide, page 68)
Is this true? Is it safe? How would I turn off the locking, and is
there a performance benefit for doing it?
Here is the full quote that you are referring to (for the record):
4.2.6 Locking and Enterprise Objects Framework
Apple’s Enterprise Objects Framework (EOF) is using OPTIMISTIC
locking; a transaction is started by,
for example, a fetch and is terminated when changes are saved.
Note that this is a conceptual user transaction. This is not
referring to an EOF or database transaction.
EOF only checks the objects that are going
to be updated. This is not entirely correct; all objects that have
been accessed should be checked.
While technically true, many objects may have been fetched in several
fetches over a period of time. Only some of them are relevant to the
change the user is committing.
The user
loads a number of rows, does some calculations and stores the
result in a row. All the rows used for the
calculation may be changed, which is undetected, and the result
would be wrong.
Assuming consistent logic, the transaction that changed one of the
rows should also have updated the calculations stored in the row
being updated. Thus, they will not match the snapshot in EOF and an
EOF optimistic locking exception will be generated.
FrontBase does implement OPTIMISTIC locking. The limited check
problem with EOF (outlined above)
can be solved by allowing nested transactions on the client; start
a transaction when the user selects
objects and commit it when the user saves the changes. Actual
ROLLBACK and COMMIT should be
made available to the user.
I have _no_ idea how that could be done in EOF. EOF is single
threaded so this seems to require one EOF stack per user. That can
eat up memory. Fast.
If the server implements the locking, the locking in the EOF is
redundant, and
snapshots etc. may be turned off.
I don't recall anyway to turn snapshots off. This seems more like a
theoretical discussion than something that is actually practical to
implement.
Chuck
--
Practical WebObjects - for developers who want to increase their
overall knowledge of WebObjects or who are trying to solve specific
problems. http://www.global-village.net/products/practical_webobjects
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden