Hi,

Has anybody seen the following, and have any ideas how to resolve?

We have two developers working on an SLX Web 7.5.4 implementation. They are both using the same server. They each have a local VFS and are deploying to separate websites and app pools.

Occassionally (and this only happens when both are actively developing/viewing records etc. at the same time) bizarre errors such as these crop up:

(Usually in ENTITY.OnCreate):

Index was outside the bounds of the array. at NHibernate.Type.TypeFactory.Assemble(Object[] row, ICacheAssembler[] types, ISessionImplementor session, Object owner) at NHibernate.Cache.Entry.CacheEntry.Assemble(Object[] values, Object result, Object id, IEntityPersister persister, IInterceptor interceptor, ISessionImplementor session) at NHibernate.Cache.Entry.CacheEntry.Assemble(Object instance, Object id, IEntityPersister persister, IInterceptor interceptor, ISessionImplementor session) at NHibernate.Event.Default.DefaultLoadEventListener.AssembleCacheEntry(CacheEntry entry, Object id, IEntityPersister persister, LoadEvent event) at NHibernate.Event.Default.DefaultLoadEventListener.LoadFromSecondLevelCache(LoadEvent event, IEntityPersister persister, LoadType options) at NHibernate.Event.Default.DefaultLoadEventListener.DoLoad(LoadEvent event, IEntityPersister persister, EntityKey keyToLoad, LoadType options) at NHibernate.Event.Default.DefaultLoadEventListener.Load(LoadEvent event, IEntityPersister persister, EntityKey keyToLoad, LoadType options) at NHibernate.Event.Default.DefaultLoadEventListener.ProxyOrLoad(LoadEvent event, IEntityPersister persister, EntityKey keyToLoad, LoadType options) at NHibernate.Event.Default.DefaultLoadEventListener.OnLoad(LoadEvent event, LoadType loadType) at NHibernate.Impl.SessionImpl.FireLoad(LoadEvent event, LoadType loadType) at NHibernate.Impl.SessionImpl.Get(String entityName, Object id) at NHibernate.Impl.SessionImpl.Get(Type entityClass, Object id) at Sage.Platform.Orm.SessionScopeWrapper.Get(Type clazz, Object id) at Sage.Platform.NHibernateRepository.NHibernateRepository.GetImpl(Object id) at Sage.Platform.NHibernateRepository.NHibernateRepository`1.Sage.Platform.Repository.IRepository.Get(Object id) at Sage.Platform.EntityFactory.GetById[T](Object id) at GeneralSearchOptionsPage.Page_PreRender(Object sender, EventArgs e) in c:\inetpub\wwwroot\SlxClient\SmartParts\Options\GeneralSearchOptionsPage.ascx.cs:line 44 at System.Web.Util.CalliHelper.EventArgFunctionCaller(IntPtr fp, Object o, Object t, EventArgs e) at System.Web.Util.CalliEventHandlerDelegateProxy.Callback(Object sender, EventArgs e) at System.Web.UI.Control.OnPreRender(EventArgs e) at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Control.PreRenderRecursiveInternal() at System.Web.UI.Page.ProcessRequestMain(Boolean includeStagesBeforeAsyncPoint, Boolean includeStagesAfterAsyncPoint

The culprit is usually around the user/tream entity. The Options screen doesnt load either (this is the error above).

Any ideas? I am thinking its an NHibernate issue, or Sage.Platform connection pooling issue..

Thanks,

Nick

I am curious how you got that configuration to work at all. I was never able to get the compiling process to be user specific vs machine specific.

Each user just creates a local VFS (in the VFS explorer in App Architect). Both import from the main VFS into that local VFS, and away you go. Just change the deployment to your own virtual directory/wwwroot folder.

Oh and rememeber to change the build path in build setting for each user otherwise things can get very interesting...

I've done all that before but i believe (and maybe its changed) that the state files and several of the config files are still maintained under the service profile or the program files\saleslogix folder, such that issues you appear to be identifying still occur. If sage has fixed that great, I wasn't awair of it yet.

Specifically, any changes to an entity gets into the program files area and is global to both users.

Well we have been working this way for a long time (probably 7.2.1 onwards) and its worked fine, up until 7.5.4...The entity references in program files are only the global ones. Each user has their own build location (somewhere in Dcuments and settings\user name\appdata etc.). I have a feeling its more to do with SessionScopeWrapper() which I have used a lot of lately. And the issues are not related to development per se but rather some session problems once the code is deployed.

The config files again only contain global setting such as referenced dll (system.data.dll etc.).

We had to as we couldn't develop locally (on our laptops) for various reasons like using a different SLX version in house. When more than one user develops from the main VFS it goes mental!

Which state files do you mean?

Remember you set your own build path (per user) in App Architect-Tools\Build Settings.

I'll have to give it another go. We moved to individual images and it makes group developement a pita issue. Because of the mix of whats stored in the lfs vs vfs and the number of components under the service account, I admit I'd given up testing the group developement process. Bottom line the one issue that cripples you got to be more trouble than it was to change to images and merging.

Thanks for the info   

Is this yet another example of the new 7.5.4. SLX Caching Services rearing its head???

We don't use multiple developers against the same database, although most of our projects use multiple developers.....

Ah, how could I forget the trusty cache server? Well switching it off so far no errors! Will keep my fingers crossed

Cheers RJ.

This makes sense as you probably have multiple caches operating on the same DB.  We'll do some research here in PSG, but I think we'll start reccomending either disabling memcache entirely for development environments, or at least making sure that all developers for a given DB are sharing the same cache server.

FYI seeing some funny errors on client sites which have been resolved by switching the cache server off. Errors such as going into an account getting "the entity you have requested "IDXXXX" cannot be located" intermittently on different records. Switch of cache server and they go away (for now anyway). What performance hit am I getting by switching it off? What exactly does it cache? Its now putting thousands of errors in the event log as I have disabled it....

Thanks,

Nick

Nick,

   The cache service is a separate process that runs as a service to cache nHibernate data.  it is only limited to the amount of free memory on the machine.  switching it off moves caching back to IIS session state which is limited in memory as well as shared with all other IIS data. in addition, separate portals (sdata, mobile, web) will not always display up to date data if you turn off memcache as they will each be caching their own data.

if you're seeing a lot of errors with it on, i would suggest opening an issue with support in order to get to the bottom of the errors.

Old post, but worth the reading...

We had this problem some days ago, and since the Nhibernate cache errors are not helpfull i added this to Account details view:

Sage.Entity.Interfaces.IAccount thisAcc = this.BindingSource.Current as Sage.Entity.Interfaces.IAccount;

Sage.Platform.EntityFactory.Evict<Sage.Entity.Interfaces.IAccount>(thisAcc.Id);

This obviously is a "workarround" and i don't personally like it.

Is there a patch to this problem in 7.5.4?