Corax indexing error: Got an invalid container id: -1

Question

Corax indexing error: Got an invalid container id: -1

loraderon opened this issue 3 months ago · comments

Using RavenDB Cloud we got an error in one of our Corax indexes. It worked earlier. I changed it to Lucene and it now works again.

Unexpected exception occurred: System.InvalidOperationException: Got an invalid container id: -1
   at Voron.Data.Containers.Container.GetMutable(LowLevelTransaction llt, Int64 id) in D:\Builds\RavenDB-Stable-6.0\60012\src\Voron\Data\Containers\Container.cs:line 1019
   at Corax.Indexing.IndexWriter.AddEntriesToTerm(Span`1 tmpBuf, Int64 idInTree, Boolean isNullTerm, EntriesModifications& entries, Int64& termId) in D:\Builds\RavenDB-Stable-6.0\60012\src\Corax\Indexing\IndexWriter.cs:line 1493
   at Corax.Indexing.IndexWriter.InsertNumericFieldDoubles(Tree entriesToTermsTree, IndexedField indexedField, Span`1 tmpBuf) in D:\Builds\RavenDB-Stable-6.0\60012\src\Corax\Indexing\IndexWriter.cs:line 1795
   at Corax.Indexing.IndexWriter.Commit[TStatsScope](TStatsScope stats) in D:\Builds\RavenDB-Stable-6.0\60012\src\Corax\Indexing\IndexWriter.cs:line 863
   at Raven.Server.Documents.Indexes.Persistence.Corax.CoraxIndexWriteOperation.Commit(IndexingStatsScope stats) in D:\Builds\RavenDB-Stable-6.0\60012\src\Raven.Server\Documents\Indexes\Persistence\Corax\CoraxIndexWriteOperation.cs:line 79
   at Raven.Server.Documents.Indexes.Index.DoIndexingWork(IndexingStatsScope stats, CancellationToken cancellationToken) in D:\Builds\RavenDB-Stable-6.0\60012\src\Raven.Server\Documents\Indexes\Index.cs:line 2478
   at Raven.Server.Documents.Indexes.Index.ExecuteIndexing() in D:\Builds\RavenDB-Stable-6.0\60012\src\Raven.Server\Documents\Indexes\Index.cs:line 1682

Arkadiusz Paliński · Answer 1 · Wed Mar 13 2024 19:28:28 GMT+0800 (China Standard Time)

@loraderon we are aware of that and have a lead already.

Maciej Aszyk · Answer 2 · Wed Mar 13 2024 19:35:11 GMT+0800 (China Standard Time)

@loraderon I created a test and issue for this.
You can track the progress here: https://issues.hibernatingrhinos.com/issue/RavenDB-22141

Arkadiusz Paliński · Answer 3 · Fri Mar 15 2024 21:17:05 GMT+0800 (China Standard Time)

This got fixed in #18240 and it will be released in stable 6.0 build next week.

Anders Wåglund · Answer 4 · Fri Mar 15 2024 23:49:31 GMT+0800 (China Standard Time)

This got fixed in #18240 and it will be released in stable 6.0 build next week.

Thank you!

Will the cloud instances also be updated next week?

Arkadiusz Paliński · Answer 5 · Mon Mar 18 2024 15:45:46 GMT+0800 (China Standard Time)

Please contact us via the support form in the cloud portal and we'll update your instance right after publishing the new build.

Arkadiusz Paliński · Answer 6 · Tue Mar 19 2024 17:38:47 GMT+0800 (China Standard Time)

This is fixed in 6.0.101. Please contact us via RavenDB Cloud portal and we'll schedule the update of your product. Please note that the reset of the affected index will be necessary after the upgrade.