[jira] [Commented] (IOTDB-1285) Support +,-,*,/ Op for SELECT syntax,e.g.: m1 + m2
[ https://issues.apache.org/jira/browse/IOTDB-1285?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17317591#comment-17317591 ] huganghui commented on IOTDB-1285: -- I am willing to try it. > Support +,-,*,/ Op for SELECT syntax,e.g.: m1 + m2 > -- > > Key: IOTDB-1285 > URL: https://issues.apache.org/jira/browse/IOTDB-1285 > Project: Apache IoTDB > Issue Type: New Feature > Components: JDBC >Reporter: sunjincheng >Priority: Major > > IoTDB> SELECT timestamp, m1 + m2 FROM root.lemming.device1; > Msg: 401: Error occurred while parsing SQL to physical plan: line 1:21 > mismatched input '+' expecting {FROM, ',', '*', '.'} -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (IOTDB-1292) Differentiable MTree
[ https://issues.apache.org/jira/browse/IOTDB-1292?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17317565#comment-17317565 ] Houliang Qi commented on IOTDB-1292: +1, It is expected to improve the performance greatly > Differentiable MTree > > > Key: IOTDB-1292 > URL: https://issues.apache.org/jira/browse/IOTDB-1292 > Project: Apache IoTDB > Issue Type: Improvement > Components: Core/Cluster >Reporter: Tian Jiang >Priority: Major > Labels: metadata, synchronization > > We have the sync tool for incremental data file synchronization, but it is > non-trivial to synchronize metadata incrementally since two MTrees cannot be > compared at a low cost. > So differentiable MTree is necessary to make it possible to compare two > MTrees and find the difference between them. When synchronizing the metadata, > only the difference will be transmitted and thus the bandwidth usage is > reduced. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Created] (IOTDB-1292) Differentiable MTree
Tian Jiang created IOTDB-1292: - Summary: Differentiable MTree Key: IOTDB-1292 URL: https://issues.apache.org/jira/browse/IOTDB-1292 Project: Apache IoTDB Issue Type: Improvement Components: Core/Cluster Reporter: Tian Jiang We have the sync tool for incremental data file synchronization, but it is non-trivial to synchronize metadata incrementally since two MTrees cannot be compared at a low cost. So differentiable MTree is necessary to make it possible to compare two MTrees and find the difference between them. When synchronizing the metadata, only the difference will be transmitted and thus the bandwidth usage is reduced. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Created] (IOTDB-1291) A potential bug in memory control of write process
Jialin Qiao created IOTDB-1291: -- Summary: A potential bug in memory control of write process Key: IOTDB-1291 URL: https://issues.apache.org/jira/browse/IOTDB-1291 Project: Apache IoTDB Issue Type: Bug Components: Core/Engine Affects Versions: 0.11.2 Reporter: Jialin Qiao Attachments: image-2021-04-08-15-59-56-983.png, image-2021-04-08-16-00-05-329.png, image-2021-04-08-16-00-16-141.png A case: Allocate 1G memory for IoTDB. 1 storage group, 1072 time series. !image-2021-04-08-15-59-56-983.png! !image-2021-04-08-16-00-05-329.png! -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Created] (IOTDB-1290) Change Inplace unseq Compaction to full compaction
Jialin Qiao created IOTDB-1290: -- Summary: Change Inplace unseq Compaction to full compaction Key: IOTDB-1290 URL: https://issues.apache.org/jira/browse/IOTDB-1290 Project: Apache IoTDB Issue Type: Improvement Components: Compaction Reporter: Jialin Qiao In the unseq compaction, we use in-place compaction to accelerate the process. However, this may generate a very large TsFile (we rewrite the overlapped Chunk behind the sequence file) There are some ways to avoid this: (1)Clean the TsFile periodically. (2)If the redundant data in a TsFile is too large, using full compaction instead of in-place compaciton -- This message was sent by Atlassian Jira (v8.3.4#803005)