The engineers at google are very vigilant at avoiding activities that cause IE to leak, so my first inclination would be to first look at your third party libs particularly if they perform any manual DOM manipulation or JSNI. It is rather common on this list to see people blaming GWT for some bug that later turns out to actually be behavior of some third party lib.
See if you can't simplify your UI (for testing purposes) so that no third party libs are needed, and then see what happens with your memory usage -jason On Feb 24, 2009, at 2:31 AM, koalina wrote: > > I've tried reloading, I made a mistake some posts ago.. memory > decreases only if the browser is closed and then opened again. Only > reloading causes a bigger increment of memory usage. > > On 24 Feb, 10:20, koalina <[email protected]> wrote: >> Hi alex, >> sorry but I can't see how fix the number of rows as I'm using a >> memoryProxy object. >> So, the timer is defined as follows: >> Timer t = new Timer() { >> >> public void run() { >> getShipInfo(); >> >> } >> >> }; >> where getShipInfo() is >> >> private void getShipInfo() { >> >> MainEntryPoint.getServiceWsNaviInMare().getInfoNaviInMare(new >> AsyncCallback() { >> >> public void onFailure(Throwable caught) { >> throw new UnsupportedOperationException("Not >> supported >> yet."); >> } >> >> public void onSuccess(Object result) { >> Vector listaNavi = (Vector) result; >> GWT.log("Lista Navi Ricevuta: "+listaNavi.toString(), >> null); >> if (listaNavi.size()>0) { >> >> Object[][] data = new Object[listaNavi.size()] >> [15]; >> Iterator i = listaNavi.iterator(); >> int index = 0; >> >> while (i.hasNext()) { >> Nave unaNave = (Nave) i.next(); >> data[index] = new Object[]{ >> unaNave.getIdNave(), unaNave.getNomeNave >> (), >> >> unaNave.getLocalita(), unaNave.getImbarco >> (), >> >> etc.... >> >> GWT.log("Nave "+data[index][0]+data[index] >> [1].toString(), null); >> >> index++; >> unaNave = null; >> >> } >> if (store!=null) { >> store.removeAll(); >> } >> MemoryProxy proxy = new MemoryProxy(data); >> ArrayReader reader = new ArrayReader(recordDef); >> store = new Store(proxy, reader); >> store.load(); >> store.commitChanges(); >> >> grid.reconfigure(store, columnModel); >> //240209Miki: x diminuire memory leak ... >> grid.clear(); >> data = null; >> proxy = null; >> reader =null; >> store = null; >> >> } >> } >> >> }); >> >> } >> >> I've tried with FF and there's no problem...so it's an issue with >> ie7. >> Can you see some big error in the code i posted just above? thx for >> help! >> Michela >> >> On 24 Feb, 09:48, "alex.d" <[email protected]> wrote: >> >>> Assuming adding/removing markers to/from the map is leakless (google >>> probably knows how to do this ;-) we are left with gwt-ext grid. >>> Every >>> 10 sec you remove old rows and add new ones. So you have some DOM- >>> Elements (a table row) that are removed/created every 10 seconds. >>> Probably it's a grid's particular implementation that leaks. But >>> most >>> probably it's just IE that doesn't really remove garbage properly >>> even >>> when it can/should. So what can you do? >>> 1. If it's grid that leaks: dump it - implement smth. yourself with >>> vanilla GWT. >>> 2. It it's IE: let's say the total ammount of markers is never >>> bigger >>> than 25 - you can create 25 rows in the table and instead of adding/ >>> removing them you just change the text. This way you'll have >>> slightly >>> more memory allocated at the beginning but hopefully no leaks while >>> running because no DOM-Elements are created/removed. >> >>> hth >> >>> On 24 Feb., 09:14, koalina <[email protected]> wrote: >> >>>> Hi Jason, yes, reloading page cause memory usage fall out. >>>> My app is quite simple: a gwt ext tab panel, the first tab >>>> showing the >>>> map widget, the second one a gwt ext grid with a few rows (more or >>>> less 10), each one with some information about the markers showed. >>>> I've three "thread", implemented as Timer, which refresh >>>> positions of >>>> markers, informations and last update time. Each thread make a rpc >>>> call and is always running, even when the tab of interest is not >>>> active.... >>>> Do you see some critical issue in this application? >>>> thx for help, have a nice day >>>> Michela >> >>>> On 23 Feb, 20:55, Jason Essington <[email protected]> >>>> wrote: >> >>>>> Well, I've read various articles claiming that IE leaks like a >>>>> sieve >>>>> (don't have any references for you right now, but I'm sure google >>>>> could help). >> >>>>> One thing to try is does IE's memory usage come down when you >>>>> reload >>>>> the page, or load a new page? >> >>>>> -jason >> >>>>> On Feb 23, 2009, at 8:44 AM, koalina wrote: >> >>>>>> is there anyone who knows about memory leaks in this case?I've a >>>>>> tabPanel, but tab are not dinamically added, a MapWidget, >>>>>> periodically >>>>>> rpc to refresh marker on map >>>>>> I'm really getting crazy! >>>>>> thx, >>>>>> Michela >> >>>>>> On 23 Feb, 14:37, koalina <[email protected]> wrote: >>>>>>> I'm using google maps api for gwt, no memory leak found >>>>>>> gwt-ext, some leaks found, but it seems not to be my case... >> >>>>>>> please note that with JScript leaks detect there's no leak >>>>>>> found... >>>>>>> could it be a problem with ajaxLoader?or perhaps something >>>>>>> coming >>>>>>> from >>>>>>> timer.scheduleRepeating with rpc inside run body? >>>>>>> Michela >> >>>>>>> On 23 Feb, 14:29, Mahavir Jain <[email protected]> wrote: >> >>>>>>>> You should try googling out for memory-leaks for the packages >>>>>>>> you >>>>>>>> are >>>>>>>> using.. >> >>>>>>>> For eg:- If you are using Image bundles then try googling the >>>>>>>> memory leaks >>>>>>>> related to image bundles.. >> >>>>>>>> Hope this will help. >> >>>>>>>> Thanks. >>>>>>>> Mahavir >> >>>>>>>> On Mon, Feb 23, 2009 at 6:49 PM, koalina <[email protected]> >>>>>>>> wrote: >> >>>>>>>>> it is present web mode running on dedicated tomcat (I don't >>>>>>>>> know >>>>>>>>> for >>>>>>>>> hosted mode..). >>>>>>>>> iexplorer.exe process exponential grows.. >> >>>>>>>>> On 23 Feb, 13:03, Eric Ayers <[email protected]> wrote: >>>>>>>>>> Is this a problem in web mode, hosted mode, or both? >> >>>>>>>>>> On Mon, Feb 23, 2009 at 7:01 AM, koalina >>>>>>>>>> <[email protected]> >>>>>>>>>> wrote: >> >>>>>>>>>>> hi Mahavir , >>>>>>>>>>> thx for your reply. >>>>>>>>>>> I'm working with gwt 1.5.3 and ie7, is this issue valid >>>>>>>>>>> also in >>>>>>>>>>> this >>>>>>>>>>> case? >>>>>>>>>>> thx, >>>>>>>>>>> Michela >> >>>>>>>>>>> On 23 Feb, 12:57, Mahavir Jain <[email protected]> wrote: >>>>>>>>>>>> I am not sure whether this applies to you or not but >>>>>>>>>>>> there is >>>>>>>>>>>> memory >>>>>>>>> leak >>>>>>>>>>>> issue in IE6 browser for all RPC call.. You can check the >>>>>>>>>>>> issue >>>>>>>>> onhttp://code.google.com/p/google-web-toolkit/issues/detail? >>>>>>>>> id=1610.. >> >>>>>>>>>>>> But this is not case for firefox.. Did you tried on >>>>>>>>>>>> firefox? >> >>>>>>>>>>>> Mahavir >> >>>>>>>>>>>> On Mon, Feb 23, 2009 at 5:13 PM, koalina >>>>>>>>>>>> <[email protected]> >>>>>>>>>>>> wrote: >> >>>>>>>>>>>>> Hi all, >>>>>>>>>>>>> I've developed a gwt app using MapWidget. The markers >>>>>>>>>>>>> showed >>>>>>>>>>>>> on the >>>>>>>>>>>>> map are periodically refreshed via rpc calls. The >>>>>>>>>>>>> refresh time >>>>>>>>> period >>>>>>>>>>>>> is 10 seconds. The app correctly runs for about an hour, >>>>>>>>>>>>> then >>>>>>>>>>>>> the >>>>>>>>>>>>> memory used increase exponentially... I've installed the >>>>>>>>>>>>> js >>>>>>>>>>>>> leaks >>>>>>>>>>>>> inspector for ie7, but no leak is catched... What can i >>>>>>>>>>>>> do to >>>>>>>>> resolve >>>>>>>>>>>>> this problem? I've reviewed all code..but everything >>>>>>>>>>>>> seems to >>>>>>>>>>>>> be >>>>>>>>>>>>> correct... >>>>>>>>>>>>> Thx for helping, have a nice day, >>>>>>>>>>>>> Michela >> >>>>>>>>>> -- >>>>>>>>>> Eric Z. Ayers - GWT Team - Atlanta, GA USAhttp:// >>>>>>>>> code.google.com/webtoolkit/ > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/Google-Web-Toolkit?hl=en -~----------~----~----~----~------~----~------~--~---
