Corrupt indexes / PrefetchingBehaviour?

[Lars Corneliussen]

I'm experiencing that indexes don't get updated. When I re-touch the docs or reset the index, things are fine.

Nor indexing errors, nothing in the logs.

I*m using scripted index results on 3800. Errors in Delete Script are not handled there....

But still it should lead to some kind of log entry if it fails there, right?

Only thing I find is:

2015-11-25 14:31:39.3715;Raven.Database.Prefetching.PrefetchingBehavior;Warn;;WASTE: Discarding future work item without using it, to reduce memory usage;

_

Lars

[Lars Corneliussen]

I also see errors like this... but not around the same time

2015-11-19 18:58:42.9697;Raven.Database.Indexing.Index.Indexing;Warn;;Failed to dispose on index update trigger in Tasks/Purreliste/Active;"Microsoft.Isam.Esent.Interop.EsentWriteConflictException: Write lock failed due to outstanding write lock

   at Raven.Database.Storage.Esent.StorageActions.DocumentStorageActions.DeleteDocument(String key, Etag etag, RavenJObject& metadata, Etag& deletedETag) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Storage\Esent\StorageActions\Documents.cs:line 639

   at Raven.Database.Actions.DocumentActions.<>c__DisplayClass5f.<Delete>b__58(IStorageActionsAccessor actions) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Actions\DocumentActions.cs:line 766

   at Raven.Storage.Esent.TransactionalStorage.Batch(Action`1 action) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Storage\Esent\TransactionalStorage.cs:line 767

   at Raven.Database.Actions.DocumentActions.Delete(String key, Etag etag, TransactionInformation transactionInformation, RavenJObject& metadata, String[] participatingIds) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Actions\DocumentActions.cs:line 814

   at Raven.Database.Actions.DocumentActions.Delete(String key, Etag etag, TransactionInformation transactionInformation, String[] participatingIds) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Actions\DocumentActions.cs:line 743

   at Raven.Database.Bundles.ScriptedIndexResults.ScriptedIndexResultsIndexTrigger.Batcher.<>c__DisplayClass4.<Dispose>b__2(IStorageActionsAccessor accessor) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Bundles\ScriptedIndexResults\ScriptedIndexResultsIndexTrigger.cs:line 174

   at Raven.Storage.Esent.TransactionalStorage.Batch(Action`1 action) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Storage\Esent\TransactionalStorage.cs:line 767

   at Raven.Database.Bundles.ScriptedIndexResults.ScriptedIndexResultsIndexTrigger.Batcher.Dispose() in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Bundles\ScriptedIndexResults\ScriptedIndexResultsIndexTrigger.cs:line 181

   at Raven.Database.Extensions.EnumerableExtensions.ApplyAndIgnoreAllErrors[T](IEnumerable`1 self, Action`1 errorAction, Action`1 action) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Extensions\EnumerableExtensions.cs:line 35"

2015-11-19 19:01:28.8760;Raven.Database.Indexing.AbstractIndexingExecuter;Warn;;Task Index: 340, Keys: ............ has failed and was deleted without completing any work;"Raven.Abstractions.Exceptions.IndexDoesNotExistsException: There is no reduce index named: 340

   at Raven.Database.Storage.Esent.StorageActions.DocumentStorageActions.TouchIndexEtag(Int32 id) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Storage\Esent\StorageActions\Indexing.cs:line 331

   at Raven.Storage.Esent.TransactionalStorage.Batch(Action`1 action) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Storage\Esent\TransactionalStorage.cs:line 767

   at Raven.Database.Tasks.RemoveFromIndexTask.Execute(WorkContext context) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Tasks\RemoveFromIndexTask.cs:line 50

   at Raven.Database.Indexing.AbstractIndexingExecuter.<>c__DisplayClass7.<ExecuteTasks>b__6(IStorageActionsAccessor actions) in c:\Builds\RavenDB-Stable-3.0\Raven.Database\Indexing\AbstractIndexingExecuter.cs:line 188"

[Oren Eini (Ayende Rahien)]

What is possible is that you are failing the transaction because you are writing to the same document in the script.

That can cause an error, but this should be handled, you can see it in the stack trace: ApplyAndIgnoreAllErrors

Do you have a way for us to reproduce this behavior?

Hibernating Rhinos Ltd  

Oren Eini l CEO l Mobile: + 972-52-548-6969

Office: +972-4-622-7811 l Fax: +972-153-4-622-7811

 

--
You received this message because you are subscribed to the Google Groups "RavenDB - 2nd generation document database" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ravendb+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Lars Corneliussen]

Haven't found a way yet. It just happens sometimes and usually under load - so I guess it is some kind of race-condition...

I think I found the other bugs i posted while trying to hunt down this one.