Hello,
I am new to everything Akka. I’ve setup a POC project/app with Akka Persistence with an Oracle DB instance.
When I run the Akka app, I get the following error:
akka.persistence.typed.internal.JournalFailureException: Exception during recovery from snapshot. PersistenceId [SomeActor|some-actor-01]. ORA-00942: table or view does not exist
Can someone help understand how to have Akka P auto generate the tables? or point me to some documentation 
Thanks in advance,
Rafael
Use SchemaUtils
https://doc.akka.io/docs/akka-persistence-jdbc/current/configuration.html#database-schema
Note that it is not recommended to auto-generate tables for production environments.
Thanks @johanandren !
Once I ran the oracle-create-schemal.sql script, I kept getting the same error. Weird. So I decided to turn on DEBUG level log in logback.xml which made me realize this is the prepared statement which is causing the error:
[2021-04-07 09:49:16,897] [DEBUG] [slick.jdbc.JdbcBackend.statement] [slick.db-1] - Preparing statement: select s94.s68, s94.s64, s94.s67, s94.s65, s94.s69, s94.s66, s94.s62, s94.s63, s94.s70 from (select “meta_ser_id” as s62, “meta_ser_manifest” as s63, “sequence_number” as s64, “snapshot_ser_id” as s65, “snapshot_payload” as s66, “created” as s67, “persistence_id” as s68, “snapshot_ser_manifest” as s69, “meta_payload” as s70 from “snapshot” where “persistence_id” = ? order by “sequence_number” desc) s94 where rownum <= 1
If you compare the oracle-create-schema.sql file to the prepared statement above, it seems the create schema script is using SNAPSHOT as the name of the table/view where as the Akka app is referencing snapshot. I am wondering if the difference in case has something to do with this error?
Perhaps the “incorrect table names” mentioned by @ning-nan here have something to do with this error/issue I am having (in which case it should obviously get fixed; either in the oracle-create-schema.sql script or in the reference.conf file)
Not sure if name case has anything to do with it, can’t say I have any Oracle specific knowledge here, but, there was at least one recent oracle related fix recently (5.0.1) Oracle schema bug: wrong attribute name · Issue #502 · akka/akka-persistence-jdbc · GitHub so I’m guessing once that was fixed it worked at least for that PR author (it’s possible he didn’t use snapshots though).
Thanks again @johanandren !
Yeah, I just confirmed that the table names and columns are case sensitive in Oracle DB; I overwrote the jdbc-snashot-store.tables.snaphot.tableName field in my application.conf file and it got me passed the issue; perhaps the reference to
include "oracle-schema-overrides.conf"
in Configuration • Akka Persistence JDBC does that exactly. Now if that file would be available somewhere for future developers to avoid running into this issue.