Dashboard Traffic Visualization Shows Abnormal High Write Activity, but Multiple Tables Are Actually Archive Tables with No Write Operations

Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.

Original topic: dashboard流量可视化大量写入异常,实际上多个表都是没有写入操作的归档表 。

| username: heming

[TiDB Usage Environment] Production Environment
[TiDB Version]
6.1.2
[Reproduction Path] Operations performed that led to the issue
insert into yixintui_operate_arch.synrpt_tt_creative_v2 select * from yixintui_operate.synrpt_tt_creative where date>=‘2022-06-01’ and ‘2022-09-01’; Similar data archiving operations.
[Encountered Issue: Phenomenon and Impact]
These operations were completed on March 3rd, and on March 7th, it was discovered that a large amount of archived table data from yixintui_operate_arch.synrpt_tt_creative_v2 for the months of July and August 2022 appeared in the dashboard traffic visualization with a lot of yellow.
image
image
image
No related table SQL can be found in the slow query logs. The yixintui_operate_arch database also has no online program operations.
[Resource Configuration]
[Attachments: Screenshots/Logs/Monitoring]

| username: heming | Original post link

There are no write hotspot regions for the archive database yixintui_operate_arch.

| username: heming | Original post link

Additionally, read hotspots are all from mysql.stats_histograms

SELECT * FROM information_schema.TIDB_HOT_REGIONS WHERE TYPE = 'read' ORDER BY flow_bytes DESC LIMIT 30;

| username: tidb狂热爱好者 | Original post link

This is optimizing the system information table, not yet completed, right?

| username: h5n1 | Original post link

I seem to recall that there is a way to clean up heatmap data.

| username: tidb狂热爱好者 | Original post link

I think the reason is that the data hasn’t been updated.

| username: tidb菜鸟一只 | Original post link

It should be the statistical information automatically collected for this data.

| username: heming | Original post link

After I backed up the data, I dropped the corresponding partitions. I’ll check the situation in the traffic visualization tomorrow.

| username: heming | Original post link

The data partitions of the archive database have all been dropped. After the tikv_gc_life_time period, everything looks normal now.

| username: system | Original post link

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.