Outbound task: Support push from Hive business email list to various databases including MySQL, jss, HBase, Oracle, jinggo, postgresql, ElasticSearch, jimdb, phoenix, etc.
Data zipper task: It supports processing the flow meter of the bdm layer into the zipper table of the fdm layer.
Data synchronization (JDW to Jmart): Support to synchronize data from data warehouse to data mart.
Through the task scheduling system, you can manage timed tasks conveniently and quickly, support the establishment of dependencies between tasks, fast replenishment and rerunning of tasks, and powerful monitoring functions, providing good job management services.
Task scheduling guarantees various tasks of 618 with strong technical capabilities, so how does a product manager of scheduling guarantee 618?
1. Beforehand: formulate a promotion strategy & publicity, and implement resource inclination
Preparatory work 1: Formulate task level classification norms, grading guarantee mechanism and management and control norms
Divide the task level into: level 0, level 1, level 2, level 3. Level 0: The company's core business, the data object is external customers or internal VPs, first-level department leaders and above. Once unavailability occurs, it will directly affect the cooperation projects of external customers, and may cause P0-P2 accidents.
Level 1: The data is oriented to the leaders of the second-level department. Once it is unavailable, it will affect the cooperation projects across the first-level department or above, and may cause the occurrence of general accidents (P3 level).
Level 2: The data object-oriented is the leader of the third-level department. Once unavailable, it will affect the internal projects of the second-level department.
Level 3: The data is oriented to the interior of the third-level department. Once unavailable, it will affect the internal project or personal report data of the third-level department. The scheduling platform will allocate resources according to the set level