[ 
https://issues.apache.org/jira/browse/IOTDB-2049?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17465570#comment-17465570
 ] 

ZhaoXin commented on IOTDB-2049:
--------------------------------

Implementation of SchemaFIle can be organized as 3 steps, listed from top to 
bottom as belowed:
 # SchemaFile Interface. It provides persistent methods for MNode from 
MTreeCache which is a part of memory-based index for IoTDB.
 # File Space Management. It records free space of SchemaFile and allocate 
pages for request from SchemaFile Interface.
 # SlottedFile Interface. It writes serialized MNode to designated Segment 
offset, with pre-allocated offset if needed. It deserialize MNode from 
designated Segment and specified node name. Page is the granularity of its 
access to fisk.

> Persistence Schema: Disk file format
> ------------------------------------
>
>                 Key: IOTDB-2049
>                 URL: https://issues.apache.org/jira/browse/IOTDB-2049
>             Project: Apache IoTDB
>          Issue Type: Improvement
>            Reporter: Jialin Qiao
>            Assignee: ZhaoXin
>            Priority: Major
>   Original Estimate: 128h
>  Remaining Estimate: 128h
>
> # Implement MetaFile I/O programs, including MNode serialization and 
> deserialization, byteBuffer read and write. 
>  # Integrate MetaFile with MNodeAccess framework to validate and test. Since 
> this, the workspace of metadata has already been extended to disk.
>  # Performance test.
>  # Implement recovery function. 



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

Reply via email to