...
Here we propose the speculative execution strategy [FLINK-10644] to to handle the problem. The basic idea is to run a copy of task on another node when the original task is identified to be long tail. The speculative task is executed in parallel with the original one and share the same failure retry mechanism. Once either task complete, the scheduler admits its output as the final result and cancel the other running one. I will introduce a blacklist module to schedule the long tail task on different machine from the original task. And modify FileOutputFormat.java to adapter speculative execution mechanism.
...
Blacklist is a kind of scheduling constraint. According to Jira server ASF JIRA serverId 5aa69414-a9e9-3523-82ec-879b028fb15b
...
I will implement (Job, Host) blacklist for speculative execution feature. In order to implement Jira server ASF JIRA serverId 5aa69414-a9e9-3523-82ec-879b028fb15b
...