On-Demand Recordings from PrestoCon’s, Webinars, Meetups, and more
We would like to talk about disaggregated coordinator production rollout with some stats. Also we can inform community about changes needed to support deployment of disaggregated coordinator.
In the existing Presto architecture, single coordinator has become a bottleneck in a number of ways for cluster scalability. – With an increasing number of workers, the coordinator has the potential of slow down due to a high number of tasks. – In high QPS use cases, we have found workers can become starved of splits by excessive CPU being spend on task updates in coordinator. – Also with single coordinator, we have an upper limit on the worker pool because of above-mentioned reasons. To overcome with this challenges, we are coming up with a new architecture which supports multiple coordinators in a single cluster.
Presto has been widely used at Bytedance in several ways such as in the data warehouse, BI tools, ads etc. And, the Presto team at Bytedance has also delivered many key features and optimizations such as the Hive UDF wrapper, coordinator, runtime filter and so on which extend Presto usages and enhance Presto stabilities. Nowadays, most companies will use both Hive (or Spark) and Presto together. But Presto UDFs have very different syntax and internal mechanisms compared with Hive UDFs. This restricts Presto usage while users need to maintain 2 kinds of functions. In this talk, we will present a way to execute Hive UDF/UDAF inside Presto.