NPE when recovering from snapshot in StreamProcessor

[npepinpe]

Describe the bug

In one case with 8.2.0-rc1, a stream processor instance failed in STARTED phase while trying to recover from snapshot, because its logStreamReader instance was null. Looking at the logs, we can see previously that it had failed to retrieve a logStreamReader instance in its STARTING phase.

We would expect the actor to have failed then, in the starting phase, but it seems it kept going into the STARTED phase, without a reader, resulting in the NPE.

See https://console.cloud.google.com/errors/detail/CJ60wpTCkMbM4AE;service=zeebe;version=release-8-2-0-rc1?project=zeebe-io

Expected behavior

No NPE occurs, and we fail the actor in the starting phase instead of moving on to the started phase when we're unable to get a log stream reader.

Log/Stacktrace

Full Stacktrace

``` ava.lang.NullPointerException: Cannot invoke "io.camunda.zeebe.logstreams.log.LogStreamReader.seekToNextEvent(long)" because "this.logStreamReader" is null at io.camunda.zeebe.stream.impl.StreamProcessor.recoverFromSnapshot(StreamProcessor.java:389) ~[zeebe-stream-platform-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] at io.camunda.zeebe.stream.impl.StreamProcessor.onActorStarted(StreamProcessor.java:161) ~[zeebe-stream-platform-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] at io.camunda.zeebe.scheduler.ActorJob.invoke(ActorJob.java:92) [zeebe-scheduler-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] at io.camunda.zeebe.scheduler.ActorJob.execute(ActorJob.java:45) [zeebe-scheduler-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] at io.camunda.zeebe.scheduler.ActorTask.execute(ActorTask.java:119) [zeebe-scheduler-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] at io.camunda.zeebe.scheduler.ActorThread.executeCurrentTask(ActorThread.java:106) [zeebe-scheduler-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] at io.camunda.zeebe.scheduler.ActorThread.doWork(ActorThread.java:87) [zeebe-scheduler-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] at io.camunda.zeebe.scheduler.ActorThread.run(ActorThread.java:198) [zeebe-scheduler-8.2.0-SNAPSHOT.jar:8.2.0-SNAPSHOT] ```

Environment:

• OS:
• Zeebe Version: 8.2.0-RC1
• Configuration:

[megglos]

ZDP Triage:

• the weird behavior was that Zeebe still entered the STARTED phase => that would make it a more severe issue
• we should verify if there a bug in the actor scheduler