You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@iotdb.apache.org by "Jialin Qiao (Jira)" <ji...@apache.org> on 2023/01/12 09:17:00 UTC
[jira] [Assigned] (IOTDB-5397) The cluster goes wrong after insert 200 records of aligned 50w timeseries
[ https://issues.apache.org/jira/browse/IOTDB-5397?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jialin Qiao reassigned IOTDB-5397:
----------------------------------
Assignee: (was: Jialin Qiao)
> The cluster goes wrong after insert 200 records of aligned 50w timeseries
> --------------------------------------------------------------------------
>
> Key: IOTDB-5397
> URL: https://issues.apache.org/jira/browse/IOTDB-5397
> Project: Apache IoTDB
> Issue Type: Bug
> Reporter: changxue
> Priority: Major
> Attachments: IOTDB-5397_allnodes-log.tar.gz, config.properties, image-2023-01-11-11-41-41-506.png, image-2023-01-11-11-42-04-777.png, image-2023-01-11-14-14-12-596.png
>
>
> The cluster goes wrong after insert 200 records of aligned 50w timeseries
> iotdb master Build: 4a90dbc
> reproduction:
> 1. start 3C3D IoTDB cluster
> 2. Insert values into IoTDB with benchmark, using the config in attachment
> actual result:
> 1. schema create successfully, value insert successfully
> 2. data node show logs :
> node 45:
> !image-2023-01-11-11-41-41-506.png|width=700!
> node 46:
> !image-2023-01-11-11-42-04-777.png|width=700!
> node 44:
> !image-2023-01-11-14-14-12-596.png|width=700!
>
> expect result:
> The IoTDB service should always work properly with normal operation.
> 线索:
> device=1
> sensors=50w
> batch-size=100
> thrift 请求: 500000*1*100*16 ~ 800M
> 谭新宇:
> 一般不建议这样来写入,需要控制每个 RPC 的大小,最好别超过几十 M
> 这里要改的话是得在客户端这边搞限制,不允许这么大的请求进来
> 最好还是能在客户端搞一些请求分批
> 问题:
> 正常的操作不应该破坏集群,使集群状态不正常
--
This message was sent by Atlassian Jira
(v8.20.10#820010)