Question

我正在尝试通过Flask服务自定义Prometheus指标。看着https://github.com/prometheus/client_python，我有一个类似于以下代码：

<svg version="1.1" baseProfile="full" width="256" height="256" xmlns="http://www.w3.org/2000/svg">
  <defs>
    <style type="text/css">
      @import url('https://fonts.googleapis.com/css?family=Dancing+Script:400');
    </style>
    <mask id="mask">
      <rect x="0" y="0" width="256" height="256" stroke-width="0" fill="white" />
      <text font-size="128" x="50%" y="50%" dominant-baseline="middle" text-anchor="middle" fill="black" style="font-family: 'Dancing Script';">T</text>
    </mask>
  </defs>
  <circle cx="128" cy="128" r="128" stroke-width="0" fill="red" mask="url(#mask)" />
  <text font-size="128" x="50%" y="50%" dominant-baseline="middle" text-anchor="middle" fill="black" style="font-family: 'Dancing Script';">T</text>
</svg>

使用此设置，我不确定是否应该在哪里声明自定义指标？

Answer 1

通过将我的自定义收集器注册到REGISTRY中解决了该问题

for c in s_values:
    df1 = FSCN_today[FSCN_today['Exporter Name'] == c]
    legal.paragraphs[7].text = c
    legal.paragraphs[8].text = df1[10:1].iloc # <- Changed

    # Add a table with the same amount of columns as the DataFrame
    table = legal.add_table(0, len(df1.columns))
    table.autofit = True

    # Create the header line (= column labels of the DataFrame)
    header = table.add_row()
    for col, cell in enumerate(header.cells):
        cell.text = str(df1.columns[col])

    # Insert the content of DataFrame in the table
    for ind in df1.index:
        row = table.add_row()
        for pos, col in enumerate(df1.columns):
            row.cells[pos].text = df1.loc[ind, col]

    # Add a break in paragraph 15 (before the table)
    legal.paragraphs[15].add_run().add_break()

    # Add the table to paragraph 15
    legal.paragraphs[15]._p.addnext(table._tbl)

    notice_name = str(c)+ ".docx"
    legal.save(notice_name)

    # Remove the table
    table._element.getparent().remove(table._element)

然后使用

REGISTRY.register(CustomCollector())

使用Flask的prometheus_client自定义指标

1 个答案: