Serialized serializer classnames non-deterministically corrupted

[beckyconning]

"org.mapdb.Serializer#BYTE_ARRA‡" note the "missing" last character.

\x6f\x72\x67\x2e\x6d\x61\x70\x64\x62\x2e\x53\x65\x72\x69\x61\x6c\x69\x7a\x65\x72\x23\x42\x59\x54\x45\x5f\x41\x52\x52\x41\xc2\x87

exception: Could not load object: org.mapdb.Serializer#BYTE_ARRA‡
List(org.mapdb.DB.nameCatalogGetClass(DB.kt:409)
, org.mapdb.DB$HashMapMaker.open2(DB.kt:879)
, org.mapdb.DB$HashMapMaker.open2(DB.kt:559)
, org.mapdb.DB$Maker.make2(DB.kt:1395)
, org.mapdb.DB$Maker.createOrOpen(DB.kt:1363)
, org.mapdb.DB$HashMapMaker.createOrOpen(DB.kt:961)
, sdbe.run.PrecogStore$.$anonfun$timestampedStore$1(PrecogStore.scala:439)
, cats.effect.internals.IORunLoop$.cats$effect$internals$IORunLoop$$loop(IORunLoop.scala:87)
, cats.effect.internals.IORunLoop$RestartCallback.signal(IORunLoop.scala:359)
, cats.effect.internals.IORunLoop$RestartCallback.apply(IORunLoop.scala:380)
, cats.effect.internals.IORunLoop$RestartCallback.apply(IORunLoop.scala:323)
, cats.effect.internals.IOShift$Tick.run(IOShift.scala:35)
, java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
, java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
, java.lang.Thread.run(Thread.java:748)
)

Happy to provide more information on request. Thanks.

[beckyconning]

This seems to be happening almost every time currently. Any idea why?

[beckyconning]

Hi, today we started work on transitioning away from MapDB due to this error affecting about a quarter of our customers.

We have read through all the MapDB source code and haven't find the offending code but if we can get some hints or tasks on what to do to resolve this we'd be happy to put the work in.

In testing we are experiencing this almost every day.

If it can be resolved we can save about a month of work and stay on MapDB.

Have you ever experienced this issue or have any idea as to what may be causing it?

We've been using MapDB 3.0.8.

Thanks!

[beckyconning]

Hi! Just to let you know we've moved away from using MapDB due to this issue. If this gets resolved in the future we may consider switching back. Thanks!