主页 / unix / 问题 / 783219
Accepted
Scott Hollenbeck
Asked: 2024-09-09 06:24:05 +0800 CST

使用 awk 解析和处理“resolvectl statistics”输出

  • 772

我正在尝试解析并重写“resolvectl statistics”命令的输出，希望生成一个 munin 插件来监控 resolvectl 处理。输出如下所示：

Transactions
                       Current Transactions:     0
                         Total Transactions: 65277

Cache
                         Current Cache Size:    13
                                 Cache Hits: 20422
                               Cache Misses: 70974

Failure Transactions
                             Total Timeouts:    32
         Total Timeouts (Stale Data Served):     0
                    Total Failure Responses:  5203
Total Failure Responses (Stale Data Served):     0

DNSSEC Verdicts
                                     Secure:     0
                                   Insecure:     0
                                      Bogus:     0
                              Indeterminate:     0

我需要重写此输出以产生如下所示的内容：

current-transactions.value 0
total-transactions.value 65277
cache-size.value 13
cache-hits.value 20422
cache-misses.value 70974
total-timeouts.value 32
total-timeouts-stale.value 0
total-failures.value 5203
total-failures-stale.value 0
dnssec-secure.value 0
dnssec-insecure.value 0
dnssec-bogus.value 0
dnssec-indeterminate.value 0

我已经能够使用这个 awk 脚本隔离这些值：

$ sudo resolvectl statistics|awk -F ":" {'print $2'}|awk NF
     0
 65227
    13
 20422
 70974
    32
     0
  5203
     0
     0
     0
     0
     0

我如何改变 awk 输出以添加到上面描述的标签？

awk

3 个回答

  1. Best Answer

    假设/理解：

    • 没有单一的规则可以确定所有标签的文本
    • 所有字符串（与值相关）都是唯一的；为了便于讨论，我将这些字符串称为“测量值”

    一般方法：

    • 预先构建一个测量字符串数组及其相关标签
    • 拆分输入行:
    • 如果有 2 个字段（NF==2）那么...
    • 删除两个字段的前导空格
    • 使用第一个字段（$1）在数组中查找相关标签
    • 将所需的标签和值打印到标准输出

    模拟OP的resolvectl输出：

    $ cat resolve.out
Transactions
                       Current Transactions:     0
                         Total Transactions: 65277

Cache
                         Current Cache Size:    13
                                 Cache Hits: 20422
                               Cache Misses: 70974

Failure Transactions
                             Total Timeouts:    32
         Total Timeouts (Stale Data Served):     0
                    Total Failure Responses:  5203
Total Failure Responses (Stale Data Served):     0

DNSSEC Verdicts
                                     Secure:     0
                                   Insecure:     0
                                      Bogus:     0
                              Indeterminate:     0

    一种awk方法：

    $ cat resolve.awk
BEGIN { FS = ":"
        labels["Current Transactions"] = "current-transactions"
        labels["Total Transactions"]   = "total-transactions"
        labels["Current Cache Size"]   = "cache_size"
        labels["Cache Hits"]           = "cache_hits"
        labels["Cache Misses"]         = "cache-misses" 

        # I will leave it up to OP to fill in the rest of the array
      }

NF==2 { sub(/^[[:space:]]+/,"",$1)          # strip leading white space from 1st field
        sub(/^[[:space:]]+/,"",$2)          # strip leading white space from 2nd field
        print labels[$1] ".value", $2       # use $1 to look up label from array and generate desired output
      }

    进行试驾：

    $ cat resolve.out | awk -f resolve.awk
current-transactions.value 0
total-transactions.value 65277
cache_size.value 13
cache_hits.value 20422
cache-misses.value 70974

    笔记：

    • OP 将替换cat resolve.out |为sudo resolvectl statistics |
    • 输出被截断为前 5 行，因为我只填充了数组中的前 5 个labels[]条目
    • 5

  2. 使用任何 awk：

    $ cat tst.awk
{ $0 = tolower($0); sub(/:/,"") }
/[^0-9]$/         { type = $1; next }
type == "cache"   { sub(/current /, "") }
type == "failure" { sub(/ response/, "") }
NF == 2           { $0 = type " " $0 }
/stale/           { $2 = $2 "-stale" }
NF { print $1 "-" $2 ".value", $NF }


    $ awk -f tst.awk file
current-transactions.value 0
total-transactions.value 65277
cache-size.value 13
cache-hits.value 20422
cache-misses.value 70974
total-timeouts.value 32
total-timeouts-stale.value 0
total-failures.value 5203
total-failures-stale.value 0
dnssec-secure.value 0
dnssec-insecure.value 0
dnssec-bogus.value 0
dnssec-indeterminate.value 0
    • 5

  3. 我会用perl：

    perl -lne '
  if (/(\S.*):\s*(\d+)/) {
    ($_, $value) = (lc($1), $2);
    y/ /-/;
    s/\(stale-data-served\)/stale/;
    s/current-(?=cache-size)//;
    s/-response//;
    s/^/dnssec-/ if $section =~ /DNSSEC/;
    print "$_.value $value";
  } else {$section = $_ if /\S/}'
    • 3

相关问题