When you query for all users in batch, the system is easily overloaded. This is the worst case query situation where no caching applies (for instance).
1) run batch queries at low input load time, because you are competing with input for access to HBase 2) throttle your query speed and/or number of connections. Hopefully you are making multiple connections in parallel. 3) scale up your HBase & Elasticsearch deployments, which tend to reach query performance limits before other services. On Oct 6, 2017, at 12:13 AM, Mattz <eatingka...@gmail.com> wrote: Hello, We are using UR with PIO for generating recommendations. We run a batch program that concurrently queries the PIO server, generates recommendations and stores in a table. Currently, we are seeing the below error several times in our logs. Any ideas on what needs to be turned to fix this? ERROR com.hifx.URAlgorithm [ForkJoinPool-4-worker-5] - Timeout when read recent events. Empty list is used. java.util.concurrent.TimeoutException: Futures timed out after [200 milliseconds] Thank you!
