[GitHub] [iotdb] wangchao316 removed a comment on pull request #2597: [IOTDB-1140] [IOTDB-1125] optimize regular data encoding

[GitBox <gi...@apache.org>]

wangchao316 removed a comment on pull request #2597:
URL: https://github.com/apache/iotdb/pull/2597#issuecomment-769614149


   > Hi, I am confused by your description...
   
   定频编码要求数据为定频数据，存在几种情况：
   1. 一次memtable 中数据不是定频数据，在flush的时候encoding会报错，直接导致本次flush的所有序列都会失败。（这是我测试的问题，本PR已解决，解决方法：将编码错误的数据过滤掉，不影响其他序列flush）
   2. 单次memtable是定频的，但是两个tsfile文件中加起来的数据不是定频的，经过合并后，会导致在读取数据解码的时候，发生错误。（本PR已解决）（这是iotdb-1125中的问题）
   3. 如果一旦flush失败，会设置成只读模式，没有恢复机制，我提了一个优化单（IOTDB-1136），将在这个单子里面解决。
   Constant-frequency encoding requires constant-frequency data. There are several cases:
   1. The data in the memtable is not constant-frequency data. During flushing, an error is reported during encoding. As a result, all sequences fail to be flushed. (This is a problem in the test and has been resolved in this PR.)
   2. A single memtable is fixed-frequency, but the sum of data in the two TSfile files is not fixed-frequency. After the combination, an error occurs during data decoding. (This issue has been resolved in this PR.) (This issue is in iotdb-1125.)
   3. If the flush fails, the system sets the data to the read-only mode and does not have a recovery mechanism. I have submitted an optimization ticket (IOTDB-1136) to resolve this issue.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org