sqlite3.OperationalError：无法打开数据库文件 - 一段时间后发生错误

Question

Tooling:

• Raspberry Pi 3B
• Raspbian
• BME280
• Python3
• Flask
• Sqlite3

Error code:

Traceback (most recent call last):
  File "BME280_DataCollector.py", line 65, in <module>
  File "BME280_DataCollector.py", line 45, in logData
sqlite3.OperationalError: unable to open database file

Working on Raspbian and want to store my sensor data in sqlite3 database. Somehow following error code occurs: "sqlite3.OperationalError: unable to open database file".

Firstly, I thought that I requested the database file too quickly and changed the the measurement time to minutes, but the error is still reproducible.

I looked into /tmp by df /tmp. But this file system is used by 12 % and not overloaded.

Also, I tried to give the full path and also the database all write and read permissions via chmod, but also no differents. In addition, I put the full path to the code.

Furthermore, I tried to make try and exception approaches which also weren't fruitful.

Nevertheless, I wanted to know if this failure occures at a certain number of interactions with the database. I found out that it always stopped at the 1020th interaction.

I also tried to restart the python script with a shell script but it didn't work out due to lack of experience and knowledge.

Code:

from flask import Flask,render_template,url_for,request,redirect, make_response, send_file
import random
import json
from time import sleep
from random import random
from flask import Flask, render_template, make_response
import datetime

import sqlite3
import sys

from matplotlib.backends.backend_agg import FigureCanvasAgg as FigureCanvas
from matplotlib.figure import Figure
import io
import os

import smbus2
import bme280


## FUNCTIONS ##
# get data
def getBMEdata():
    port = 1
    adress = 0x77
    bus = smbus2.SMBus(port)
    calibration_params = bme280.load_calibration_params(bus, adress)
        
    bme_data = bme280.sample(bus, adress, calibration_params)
    temp = '%.2f' % bme_data.temperature
    hum = '%.2f' % bme_data.humidity
    press = '%.2f' % bme_data.pressure
    
    now = datetime.datetime.now() #get time
    timeString = now.strftime('%d-%m-%Y %H:%M:%S') #write time to string
        
    return temp, hum, press, timeString

# function to insert data on a table
def logData(temp, hum, press, timeString):
    conn = sqlite3.connect(dbname)
    curs = conn.cursor()
    curs.execute("INSERT INTO BME280_data values((?), (?), (?), (?))", (timeString, temp, hum, press))
    conn.commit()
    conn.close()
        
# display data base
def displayData():
    conn = sqlite3.connect(dbname)
    curs = conn.cursor()
    print("\nEntire database contents:\n")
    for row in curs.execute("SELECT * FROM BME280_data"):
        print(row)
    conn.close()
    
## MAIN
if __name__ == '__main__':
    count = 0
    dbname = '/home/pi/projects/SmartPlanting/Sensors_Database/sensorsData.db'
    sampleFreq = 60 #data collect every minute
    while True:
        temp, hum, press, timeString = getBMEdata() #get data
        logData(temp, hum, press, timeString) #save data
        sleep(sampleFreq) #wait
        displayData() #show in terminal
        #count = count+1
        #print(count)

Maybe someone already solved this problem or can give me an alternative to sqlite3 which works with flask.

Answer 1

Suggestion: add a more complete exception handling routine, because your stacktrace could be more verbose.

But judging from your trace the offending line #45 could be this: conn.commit() (or the line above). Python is already helping you pinpoint the error. There is something wrong in function logData.

Could it be that you are feeding incorrect data to your table BME280_data ? To debug your application I would strongly recommend that you print log the data you are trying to insert (use the logging module to output to file and/or console). I don't know the structure of your table but some of your fields could have a definition (data type) that is not compatible with the data you are trying to insert. The fact that you are able to predictably reproduce the problem is quite telling and my hunch is that the data could be the cause.

To sum up: take good habits now and add at least basic exception handling. A quality application should have exception handling and log errors so they can be reviewed and remediated by a human. This is even more important for unattended applications, because you are not in front of the console and you may not even have a chance to see problems that occur.

Here is one tutorial that may help: https://code.tutsplus.com/tutorials/error-handling-logging-in-python--cms-27932