How to handle 409 exception while flushing buffer

We have a fluent configuration where we listen to 2 rabbitmq. we had created a separate buffer for each source. the first buffer tries to create a record with pending status. The second buffer tries to update the record with status as success/failure. please find buffer configuration.

    @type file
    path /fluentd/buffer
    flush_interval 5s
    flush_mode interval
    retry_wait 3                      # The wait interval for the first retry.
    retry_exponential_backoff_base 2  # Inclease the wait time by a factor of N.
    retry_type exponential_backoff    # Set 'periodic' for constant intervals.
    retry_max_interval 1h             # Cap the wait interval. (see above)
    retry_randomize true              # Apply randomization. (see above)
    retry_timeout 24h                 # Maximum duration before giving up.
    retry_forever false               # Set 'true' for infinite retry loops.
    retry_secondary_threshold 0.8     # See the "Secondary Output" section in
    flush_thread_count 3
    overflow_action block

When flush is happening to elastic search in a bulk request, the response is success but internally one record update is failure

{"update"=>{"_index"=>"cars-2021.06.09", "_type"=>"fluentd", "_id"=>"1127429471072048942891864682698024", "status"=>409, "error"=>{"type"=>"version_conflict_engine_exception", "reason"=>"[fluentd][1127429471072048942891864682698024]: version conflict, document already exists (current version [1])", "index_uuid"=>"VJ9-odowS0W6bqvzRBN8qA", 
"shard"=>"0", "index"=>"cars-2021.06.09"}}},

How to retry this kind of errors which is one record failure in bulk request