Problems that occur when continuously inserting data?

[elhaz]

Question Summary:

I tried inputting my data into InfluxDB to test whether the data storage method of the existing project could be replaced with InfluxDB. (In the past, sensor data was saved as a file system.)
For the first few minutes of input, when searching with Flux query, the entire data could be searched quickly, but when data was continuously entered, it took about 4 hours for the number of points of the entered data to be searched. (After 4 hours, all data is searched.)

When entering data, it was confirmed that the hard disk capacity immediately increased regardless of the query inquiry.
Indexing the data in InfluxDB (I don't know if this term is correct), I guess it takes about 4 hours.
Can you tell me what is going on within Influx DB?

specific experimental environment

Test environment configuration:
Configured in Windows Hyper-V.
Host OS: Windows 10 Pro
Guest OS: (InfluxDB OSS 2.2) Windows 10 Pro, RAM 16GB
Guest OS: (Python 3.10.4, influxdb-client==1.26.0) Windows 10 Pro, RAM 16GB

Conditions for generating test data:
(Python Client)
now-now+4 hours range
Generate time data every 999 milliseconds
Create 2500 data column row data
csv file, 14415 lines excluding header, and the capacity is about 240 mb.

Data insertion condition:
Python client – ​​using the InfluxDBClient.write_api().write() method.
SYNCHRONOUS is used for write_option.
Record = data point.
All data input sequentially by making csv 1line a point.
When input, the specific time of the test data, not the current time, is included in the point through the point.time() method and sent.

The process of creating a csv file and inputting it to Influx with the Python client was repeated infinitely.
One file is created and input takes about 4 minutes.

Below is the Flux query I used for the query.

from(bucket: "p1")
|> range(start: -5h, )
|> filter(fn: (r) => r["_field"] == "sensor0001")
|> count()
|> yield()

This is an example of a query result.

_measualment | _value |

oFunHuJs | 14415 | << over 4 hours
5wlIxpuP | 14415 | << over 4 hours
dGCNY7kH | 14415 | << over 4 hours
lqmByzst | 14415 | << over 4 hours
rTf7iZfQ | 14415 | << over 4 hours
sP4TvxN6 | 14415 | << over 4 hours
rLZC43sD | 14415 | << over 4 hours
zVLHGwv3 | 14002 | << about 4 hours ago
lg1MiT6i | 10347 | << about 3 hours ago
eR1lrXAO | 7725 | << about 2 hours ago
MrwKdXhF | 4012 | << about 1 hour ago
8h4MfJgU | 415 | << a few minutes ago

[elhaz]

Answer yourself:
This happened because the query condition was given incorrectly.
This is because the data creation time was created up to 4 hours after the current time.
I'm very ashamed...