Junfan Zhang

[Feature] Introduce graceful decommission for shuffle server

7

### What changes were proposed in this pull request? [Feature] Introduce graceful decommission for shuffle server ### Why are the changes needed? When shuffle servers needed to decommission are included...

[Performance Optimization] The huge performance drop due to the method of getBlockIdsByPartitionId

3

### Background I found when getting shuffle result, the flame graph show the method of `getBlockIdsByPartitionId` occupy too much time.  But i found the default bitmapNum is 1, and...

[Feature Request] Introduce the unique trace id to debug easily

4

### Motivation It's hard to analyze which process cost the most time of one remote request in current codebase, due to lacking corresponding trace id in client and server side....

stockState 存在 bug, 针对有些商品有货，但是需要定时抢购的，存在直接下单，导致账号封停现象

1

明天开一个 pull request 修复

只是看代码觉得此处有问题，望指正。 如有问题，我　pull request　一下

… of Services.get().getConf().getInt

Best exporter support final best evaluation result output

5

### Why In estimator, we need to print final best eval result when using the best exporter.

Support Hadoop Compatible Filesystem interface using config instead of hard-code

4

### Why In our internal env, we develop a new filesystem implementing HCFS. To serve for users better, we hope users could use tf dataset api to read new filesystem...

Task executors that support specific roles are restarted when they fail

2

### Why Now TonY introduces the Sidecar Tensorboard, but sometimes it will fail due to hardware problems and unstable HDFS. But for users, it's better to unconscious restart it. So...

Currently, venv only can be compressed with zip format. Maybe we should support `tar.gz` format.