Conflating

[Thomas Lo]

Peter,

Would like to get your opinion on the usage of Chronicle. We are planning to use Chronicle for journaling, messaging passing, and also as a processing queue to process the incoming events. However, we are hitting one use case where we are not sure if we should utilize Chronicle to solve is conflating. Where for certain type of events (say market data of a particular product stream), if all of a sudden the volume drives up, instead of adding fuel to the fire to get the CPU to process all the event even though we only care about the latest, we would like to just conflate that type of event and just process the latest one. Have you come across that type of situation and if so how do you solve it in Chronicle's perspective. Or we should handle that outside of Chronicle? Thanks

Thomas

[Peter Lawrey]

Chronicle is designed to store and process every event.  However, if you need conflation, SharedHashMap may be a useful addition.  This is a keyed, off heap store which will give you the latest value.  You can write your own merging logic if plain value replacement is not suitable.  It will support TCP replication, however for now, I suggest using chronicle TCP replication as you are writing to chronicle already.

--
You received this message because you are subscribed to the Google Groups "Java Chronicle" group.
To unsubscribe from this group and stop receiving emails from it, send an email to java-chronicl...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Thomas Lo]

Thanks.