Pandas数据框一旦部署到heroku（Django项目）后就无法按预期方式运行

Question

我创建了一个程序，该程序可以进行数据输入并返回一个熊猫表（信息存储在views.py文件顶部的列表中，称为JOB_INFO-我将在下面附加此列表），结果如下：已下载。我还在具有数据输入表单的html页面上具有表呈现-该表呈现在表单下方，并随着每个用户输入（表单提交）进行更新（追加）。这在我的本地计算机上的开发中非常完美，但是当成功将其部署到heroku时，它就会开始起作用，并且行为异常。

示例包括：不更新表格-表单似乎已提交，但表格未更新。如果我继续尝试，它会突然起作用。但随后，我将尝试另一种，而上一个将消失，而表中将仅显示最新的一个。似乎太随意了。不清空列表-我有一个按钮可以清除表格，但这也不起作用。如果我多次单击按钮，它可能会突然工作。但是然后随机地，一些信息会再次出现在表中，我认为它已被清除！当我以为我已清除文件时，下载文件中的信息-这是不言自明的。

简而言之，将其称为随机可能是描述它的最佳方法。它只是不做我想做的事情和对它的期望。

views.py

from django.shortcuts import render, redirect, get_object_or_404, reverse, HttpResponse
from django.core.exceptions import ObjectDoesNotExist, MultipleObjectsReturned
from .models import Data
from .forms import DataForm, SearchAndUpdateForm
from io import BytesIO
import pandas as pd
import xlsxwriter


JOB_INFO = {
    "Plentific Job Number": [],
    "WorkOrder Number": [],
    "Invoice Number Found": [],
    "Payments": [],
    "Job Number Exists": [],
} 


def data_entry(request):
    """
    Creates new instance of data entry form and renders it on html page
    along with a report table, which updates with each data entry input
    """

    form = SearchAndUpdateForm()

    df = pd.DataFrame(JOB_INFO,
        columns=[
            "Plentific Job Number",
            "WorkOrder Number",
            "Invoice Number Found",
            "Payments",
            "Job Number Exists",
        ])

    if not df.empty:
        df_html = df.to_html(classes="table table-striped table-hover")
        return render(request, "data_entry.html", {
            "form": form, "df_html": df_html})
    else:
        return render(request, "data_entry.html", {
                "form": form})


def export_job_info_report(request):
    """
    Exports the job_info list as an Excel spreadsheet
    and saves it to user's local directory
    """

    df = pd.DataFrame(JOB_INFO)

    output = BytesIO()
    writer = pd.ExcelWriter(output, engine='xlsxwriter')
    df.to_excel(writer, sheet_name='report', index=False)
    writer.save()
    output.seek(0)
    response = HttpResponse(output,
        content_type='application/vnd.openxmlformats-officedocument.spreadsheetml.sheet')
    response['Content-Disposition'] = 'attachment; filename=%s.xlsx' % 'Download'

    return response


def clear_report_info(request):
    """Empties the job_info dictionary"""

    JOB_INFO["Plentific Job Number"].clear()
    JOB_INFO["WorkOrder Number"].clear()
    JOB_INFO["Invoice Number Found"].clear()
    JOB_INFO["Payments"].clear()
    JOB_INFO["Job Number Exists"].clear()

    return redirect(reverse(data_entry))

html页面

{% extends "base.html" %}
{% block content %}

{% load crispy_forms_tags %}

<div class="jumbotron">

    <h1>Data Entry Tool:</h1>

    <form action="{% url 'search_update_feedback' %}" method="GET">
        {{ form|crispy }}
        <button class="btn btn-success" type="submit" data-toggle="tooltip" data-placement="right" title="Updates the database with the info provided above">
            Update and Fetch Info
        </button>
    </form>

    <br>

    <form action="{% url 'export_job_info_report' %}" method="GET">
        <button class="btn btn-success" type="submit" data-toggle="tooltip" data-placement="right" title="Downloads the report below in .xlsx format">
            Download Report
        </button>
    </form>

    <br>

    <form action="{% url 'clear_report_info' %}" method="GET">
        <button class="btn btn-success" type="submit" data-toggle="tooltip" data-placement="right" title="Warning! You will lose all of the data below (does not affect database data)">
            Clear Report - hover over me!
        </button>
    </form>

</div>

<hr>

{{ df_html | safe }}

{% endblock %}

网址

urlpatterns = [
    path("", data_entry, name="data_entry"),
    path("export_job_info_report/", export_job_info_report, name="export_job_info_report"),
    path("clear_report_info/", clear_report_info, name="clear_report_info"),
]

当然，我省去了很多功能来节省阅读时间，但是如果可以/应该添加其他任何内容以帮助理解该问题，请告诉我。任何帮助将不胜感激。