Re: 5.3 CLOSE_WAIT problem and multi instance change notification
Re: 5.3 CLOSE_WAIT problem and multi instance change notification
- Subject: Re: 5.3 CLOSE_WAIT problem and multi instance change notification
- From: Chuck Hill <email@hidden>
- Date: Fri, 14 Oct 2005 10:34:14 -0700
On Oct 12, 2005, at 11:33 AM, Hordur Thordarson wrote:
On 12.10.2005, at 18:26, Chuck Hill wrote:
On Oct 12, 2005, at 11:18 AM, Hordur Thordarson wrote:
<snip>
If they don't communicate the change with change notification
then the only ways I can see are the ones suggested by you/Apple:
- Timed invalidation via EO(F)'s fetch timestamps
This applies only to the first fetch/fault of an object into an
editing context or subsequent refaulting of particular objects.
Hmm, sounds like I need to re-read this. I was assuming that
functioned like so given a 60 minute refetch timeout:
- An app requests an EO for the first time and it gets fetched
from the db and gets the current timestamp
- 30 minutes later the app re-requests the EO and gets the cached
version from the object graph
- 2 hrs down the road the app re-requests again and EOF then goes
to the db again as the EO is over 60 minutes old in the cache
Am I misunderstanding this ?
Yes. :-) This has _nothing_ to do with the application. It only
applies to editing context's getting a copy of the data from the
DB snapshots and asking that it receive nothing older than X.
Once the EC has that copy of the data, the age of the data is not
considered again unless the object is re-faulted / invalidated.
Creating new ECs frequently and avoiding the long lived
session.defaultEditingContext() will result in much fresher data.
<LIGHT GOES ON/>
Ok I get it now :-) I thought the other functionality sounded
pretty great ! Automatic timebased re-fetching of objects into an
EC, not only into the DB snapshots...
That would be interesting functionality, I'm not sure why it was not
done like that. Perhaps performance concerns, I have not really
thought it through.
Anyway, I see what you are saying now re. diligent EC creation
rather than storing a monster EC somewhere. If there weren't a
single, global, application EC but rather f.ex. an EC for each DA
request then this timestamping would work the way I thought it
would, right ?
Correct.
Chuck
--
Coming in 2006 - an introduction to web applications using WebObjects
and Xcode http://www.global-village.net/wointro
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