[02:33:45 CDT(-0500)] <jgribonvald> Hi, is there someone at this time ?
[02:34:46 CDT(-0500)] <jgribonvald> morning in France ...
[10:33:11 CDT(-0500)] <jgribonvald> Hi, are you interested by a feedback of our deployment in up 4.0.12 ?
[10:33:30 CDT(-0500)] <tlev> sure, how did it go?
[10:33:52 CDT(-0500)] <jgribonvald> well now after some bad things
[10:33:57 CDT(-0500)] <jgribonvald> the main one is
[10:34:21 CDT(-0500)] <jgribonvald> never deploy a JSF portlet that use the myface portlet-bridge
[10:34:34 CDT(-0500)] <jgribonvald> there is a main bug
[10:36:15 CDT(-0500)] <jgribonvald> with the bug the portal go on overload
[10:36:58 CDT(-0500)] <jgribonvald> so we solved the bug but we will merge all our old portlet using JSF to spring MVC 
[10:37:31 CDT(-0500)] <jgribonvald> after we had a concurrency problem with the user creation
[10:37:39 CDT(-0500)] <jgribonvald> with mysql
[10:37:48 CDT(-0500)] <jgribonvald> mysql created deadlock
[10:37:58 CDT(-0500)] <jgribonvald> due to our load-balancing
[10:38:09 CDT(-0500)] <tlev> ah right with the same user coming from two nodes
[10:38:18 CDT(-0500)] <jgribonvald> yes
[10:38:41 CDT(-0500)] <jgribonvald> but we think that's a problem with apache that gave bad route
[10:38:53 CDT(-0500)] <jgribonvald> after a user refreshed
[10:39:56 CDT(-0500)] <jgribonvald> we don't know why but the load-balancer have some times problems with balancer member in error during 2 or 3 seconds
[10:40:11 CDT(-0500)] <jgribonvald> and only in a higth load
[10:40:54 CDT(-0500)] <jgribonvald> but for us the problem is to watch from what this is comming, we can't place log on ajp without to rebuild debian packes
[10:41:34 CDT(-0500)] <jgribonvald> we solved this problem by adding a table lock when creating a user
[10:41:56 CDT(-0500)] <jgribonvald> and we inserted the main part of our active users
[10:42:10 CDT(-0500)] <jgribonvald> with files definitions
[10:43:00 CDT(-0500)] <jgribonvald> And now one of our next problem is external integrated link in http whereas our portail is in https ...
[10:43:39 CDT(-0500)] <jgribonvald> but for other things this going well, expect that we lost a lot of users after the problems we had
[10:44:31 CDT(-0500)] <jgribonvald> So now I have to watch on our improvment to make some PR
[10:47:24 CDT(-0500)] <jgribonvald> tlev: when watching on PR i have again 2 PR in waiting UP-3760 - Adding portlet on locked tab/column and UP-3755 - Adding Portlet Parameter 'alternativeMaximixedLink' what is your feeling about to integrate it ? in my case it's in production on our 4.0.12 version
[10:48:43 CDT(-0500)] <tlev> I will try to look at these sometime this week
[10:50:22 CDT(-0500)] <jgribonvald> ok thanks you
[10:51:18 CDT(-0500)] <jgribonvald> Also we are working with esup about an alternative smartldapGroupStore using Grouper
[10:51:44 CDT(-0500)] <jgribonvald> with grouper there is a tool to publish group in ldap
[10:51:55 CDT(-0500)] <jgribonvald> and we are using this groupStore
[10:52:53 CDT(-0500)] <jgribonvald> so this could be also a feedback
[10:53:21 CDT(-0500)] <jgribonvald> we have around 39k group maintained in memory from this group store
[10:54:24 CDT(-0500)] <jgribonvald> but we surely have to talk about some performance problems with portlets using the method isuserinRole that request massive groups