Relational agent question

[Vito]

Dear staff,
I have a strange problem with the relational agent. I noticed that when this agent sends a group of measurements (10) to the database, it writes such line in the log file:

2020-03-10 11:32:08,021 UTC [org.springframework.data.redis.listener.RedisMessageListenerContainer0-129243] DEBUG org.sentilo.agent.relational.repository.impl.AgentRelationalBatchRepositoryImpl - Scheduling batch update task for persist 10 elements in dataSource sentiloDs

This is the normal behaviour and there isn't any mystery... But sometimes, I can see a different log entry:

2020-03-10 11:33:30,378 UTC [platformScheduler-3] DEBUG org.sentilo.agent.relational.repository.impl.AgentRelationalBatchRepositoryImpl - Scheduling batch update task for persist 10 elements in dataSource sentiloDs

In such case, I think this scheduler is trying to write older (and wrong/malformed) measurements, which weren't be written on the first try. Is it possible? If yes, how can I flush/delete those measurements? Neither restarting the relational agent, nor restarting the server, fixes the problem...

Best regards,

Vito

[Vito]

Dear staff,

maybe my question isn't clear enough... So I can try to ask it in an alternative way... When I get this message:

Number of retries XXX is greater or equals than the maximum number of retries configured YYY. Events to persist will be stored in the pending queue for further processing.

Where is that queue? and how to clear it? I'm sure it's a persistent queue, since I can't clear it by restarting services or the server, but I'm not able to find it in Redis or any file...

Best regards,

Vito

[Vito]

Ok, I found it :-) it's stored in the "relational:pending:events" Redis key...

Best regards,

Vito

[Sentilo IO]

Hi Vito,

Sorry, we couldn't answer you ... but it's exactly as you say. 

These kind of agents has this feature to assure not to lose any data.

Regards,

Sentilo Team.