Kyungtae Kim

This PR resolves #1139 by : - Adding `ReplayServerAddress` and `ReplayServerPort` parameters - Modifying `start` method of `QueryStarter` class to `startOrSetupForReplay` because most of the code for both are the...

When a group is recovered, events must be replayed before the execution DAGs start receiving new events. Therefore the current implementation of the query `create` method must be separated into...

Currently, sources of a recovered query are started immediately after the single query is recovered. However, before sources of recovered queries are started, it should be ensured that all queries...

Currently, there are operators that have generics in their states. If these are not Serializable, the serialization process will inevitably fail. - `ReduceByKeyOperator`: Key/Value of `HashMap` - `ApplyStatefulOperator:` state of...

To properly recover from checkpoints, the queries must be resubmitted in the same order as the order they were submitted initially.

Currently, the MistDriver assigns a maximum number of (Total # of tasks/Total # of masters) queries to a single master. It packs each master with tasks, then moves on to...