Apr 14, 2014, 3:52 PM
[CLOSED] TreeStore-Node commit is exceptionally slow
Hi,
We are having issues with node's commit method, we have a complex screen that has 8 TreeStores bound with 8 TreeGrids. These treegrids allow editing and on each cell-edit certain calculation takes place.
These calculation spans through multiple treestores, hence we have to commit each node post calculation so that the bound tree-grid shows the changes.
We have realized that committing each node consumes lots of time, earlier we were suspecting that the find operation is eating up time but lots of debugging with messages proved that the commit operation takes most of the time (we also tried using set method but that is even slower). How can we speed this up?
What have we tried so far?
1) Wrapped the commit between suspendEvent and resumeEvent, this brought a little respite but still its pretty slow.
2) Refreshed View instead of commit so that the view reads the store and shows the result but this deteriorated the performance.
Is there a better way to approach this requirement?
Any help will be greatly appreciated!
Cheers,
Avi
We are having issues with node's commit method, we have a complex screen that has 8 TreeStores bound with 8 TreeGrids. These treegrids allow editing and on each cell-edit certain calculation takes place.
These calculation spans through multiple treestores, hence we have to commit each node post calculation so that the bound tree-grid shows the changes.
We have realized that committing each node consumes lots of time, earlier we were suspecting that the find operation is eating up time but lots of debugging with messages proved that the commit operation takes most of the time (we also tried using set method but that is even slower). How can we speed this up?
What have we tried so far?
1) Wrapped the commit between suspendEvent and resumeEvent, this brought a little respite but still its pretty slow.
2) Refreshed View instead of commit so that the view reads the store and shows the result but this deteriorated the performance.
Is there a better way to approach this requirement?
Any help will be greatly appreciated!
Cheers,
Avi
Last edited by Daniil; Apr 20, 2014 at 3:54 PM.
Reason: [CLOSED]