Details
-
Type:
Bug
-
Status: Closed
-
Priority:
Major
-
Resolution: Fixed
-
Affects Version/s: 2.0-Alpha2
-
Fix Version/s: 2.0-Alpha3, 2.0.0
-
Component/s: Bridge, Framework, Sample Apps
-
Labels:None
-
Environment:ICEfaces 2 ICEpush JSF 2
Description
Currently, when running the auction application, updates based on push (which occur every second) are causing very high CPU usage and preventing the user from bidding, chatting, etc. Looking at the updates, each one is a ViewBody update rather than just the changed clock values. It looks like it may be related to the script used for setting up ICEpush:
<script type="text/javascript">ice.setupPush('vvm2tk61', 'bkg92z89rg:se');</script>
where the window id may be changing on every push. If the clock interval is reduced significantly (e.g. 30 seconds), it reduce the impact on the app and the auction app is then usable again - although I didn't see any successful push updates after that.
<script type="text/javascript">ice.setupPush('vvm2tk61', 'bkg92z89rg:se');</script>
where the window id may be changing on every push. If the clock interval is reduced significantly (e.g. 30 seconds), it reduce the impact on the app and the auction app is then usable again - although I didn't see any successful push updates after that.
Activity
- All
- Comments
- History
- Activity
- Remote Attachments
- Subversion
The request sent for retrieving updates didn't send the ice.window parameter because of the recent changes. This eventually made the server generate window IDs for each request and in turn generate 'body' level updates because the input element holding the window ID would change every time.
The solution was to just send the window ID along.
Show
Mircea Toma
added a comment - The request sent for retrieving updates didn't send the ice.window parameter because of the recent changes. This eventually made the server generate window IDs for each request and in turn generate 'body' level updates because the input element holding the window ID would change every time.
The solution was to just send the window ID along.
This is preventing QA from doing any useful testing on glimmer thus the higher priority.