All our jobs (using streamsx.messagehub toolkit 2.0.1) are currently failing to reset from checkpoints with the following error:
at com.ibm.streamsx.kafka.clients.consumer.CrKafkaConsumerGroupClient.createSeekOffsetMap(CrKafkaConsumerGroupClient.java:1045)
at com.ibm.streamsx.kafka.clients.consumer.CrKafkaConsumerGroupClient.resetPrepareDataBeforeStopPolling(CrKafkaConsumerGroupClient.java:909)
at com.ibm.streamsx.kafka.clients.consumer.CrKafkaConsumerGroupClient.onReset(CrKafkaConsumerGroupClient.java:481)
at com.ibm.streamsx.kafka.operators.AbstractKafkaConsumerOperator.reset(AbstractKafkaConsumerOperator.java:1100)
When enabling debug tracing we get the explanation in the following error:
07 Aug 2019 08:15:44.725+0000 [49763] DEBUG M[ClientNotifForwarder.NotifFetcher-run:?:-1] - Failed to forward a notification to a listener
07 Aug 2019 08:15:44.726+0000 [49763] DEBUG M[?:?:0] - com.google.gson.JsonSyntaxException: com.google.gson.stream.MalformedJsonException: Expected ':' at line 1 column 418 path $.key.192
07 Aug 2019 08:15:44.726+0000 [49763] DEBUG M[?:?:0] - com.google.gson.Gson.fromJson(Gson.java:902)
07 Aug 2019 08:15:44.726+0000 [49763] DEBUG M[?:?:0] - com.google.gson.Gson.fromJson(Gson.java:852)
07 Aug 2019 08:15:44.727+0000 [49763] DEBUG M[?:?:0] - com.google.gson.Gson.fromJson(Gson.java:801)
07 Aug 2019 08:15:44.727+0000 [49763] DEBUG M[?:?:0] - com.google.gson.Gson.fromJson(Gson.java:773)
07 Aug 2019 08:15:44.727+0000 [49763] DEBUG M[?:?:0] - com.ibm.streamsx.kafka.clients.consumer.CrKafkaConsumerGroupClient.handleNotification(CrKafkaConsumerGroupClient.java:270)
07 Aug 2019 08:15:44.727+0000 [49763] DEBUG M[?:?:0] - com.sun.jmx.remote.opt.internal.ClientNotifForwarder$NotifFetcher.dispatchNotification(ClientNotifForwarder.java:421)
07 Aug 2019 08:15:44.728+0000 [49763] DEBUG M[?:?:0] - com.sun.jmx.remote.opt.internal.ClientNotifForwarder$NotifFetcher.run(ClientNotifForwarder.java:395)
07 Aug 2019 08:15:44.728+0000 [49763] DEBUG M[?:?:0] - java.lang.Thread.run(Thread.java:812)
Turns out when the sequence id passes 1000 the json generation is incorrectly adding number separators ( 1,000).
All our jobs (using streamsx.messagehub toolkit 2.0.1) are currently failing to reset from checkpoints with the following error:
When enabling debug tracing we get the explanation in the following error:
Turns out when the sequence id passes 1000 the json generation is incorrectly adding number separators (
1,000
).