Streaming dependency items about jaeger-analytics-java HOT 2 OPEN

I am glad to see this effort 👍🏼

I do have one question on how the results (dependency links) are stored and queried from the storage. The current spark batch implementation creates one record per time interval (e.g. 24h). I assume the streaming creates multiple records per the same interval. Did you have to change the query or perhaps the store implementation to deal with this?

from jaeger-analytics-java.

@pavolloffay no we didn't. Jaeger queries for a certain period and gets a set of dependency links (items) back. Depending on the time interval for the dependencies, the query will return a bigger or smaller set. We didn't change data models and queries. The time interval for dependency links can be set through properties.

from jaeger-analytics-java.