full list of metrics used in our sentiment analysis software

sentiment.analysis.metrics.list.txt

A complete list of the metrics and names I added were:
latency metrics list
processing_duration_ms - histogram - which measures processing latencies.
deserialize_duration_ms - histogram - which measures how long it takes to deserialize a request
serialize_duration_ms - histogram - which measures how long it takes to serialize an event
store_duration_ms - histogram - which measures how long it takes to store(index/save/persist) a state change.
with all of these, I can now define a total_processing_duration_ms which will be deserialize_duration_ms + serialize_duration_ms + store_duration_ms + processing_duration_ms

Applicative usage metrics
incoming_requests_count - counter - which will indicate how many requests did we process. Using this metric we can define throughput!
outgoing_events_count - counter - which will indicate how many events did we publish.
failed_requests_count - counter - which will indicate how many requests did we fail to process.
sentiment_score_detected- counter - which will indicate what is the score per sentiment that was calculated per every request.
sentiment_score_per_tweet - gauge - which will report what's the sentiment score per classified tweet.
sentiment_values_length -gauge - which will report how many sentiment values did we classify the tweet.
request_size_chars - histogram - which will indicate what's the size of the current request in bytes.
outgoing_event_size_chars - histogram - which will indicate what's the size of the current event size that was published in bytes.