Telegraf: Metric buffer limit: Unterschied zwischen den Versionen
Aus Wiki-WebPerfect
Admin (Diskussion | Beiträge) |
Admin (Diskussion | Beiträge) |
||
(2 dazwischenliegende Versionen des gleichen Benutzers werden nicht angezeigt) | |||
Zeile 7: | Zeile 7: | ||
<span style="color:blue">'''INTERVAL'''</span> = The average interval that Telegraf collects the metrics. <br> | <span style="color:blue">'''INTERVAL'''</span> = The average interval that Telegraf collects the metrics. <br> | ||
− | '''Calculation''': <span style="color:red">'''METRIC_BUFFER'''</span> / <span style="color:green">'''METRICS_PER_INTERVAL'''</span | + | '''Calculation''': (<span style="color:red">'''METRIC_BUFFER'''</span> / <span style="color:green">'''METRICS_PER_INTERVAL'''</span>) * <span style="color:blue">'''INTERVAL'''</span> / 60 = ''Number of minutes Telegraf buffers metric locally'' |
− | + | ||
+ | '''Example:''' (10000 / 700) * 30 / 60 = ~7 minutes | ||
Aktuelle Version vom 10. März 2021, 14:18 Uhr
The metric_buffer_limit of Telegraf is the Maximum number of unwritten metrics per output. Increasing this value allows for longer periods of output downtime without dropping metrics at the cost of higher maximum memory usage.
Calculate how much time it takes to reach the metric buffer limit
METRIC_BUFFER = Telegraf configuration "metric_buffer_limit". (Default = 10000)
METRICS_PER_INTERVAL = Run Telegraf in "Test-Mode" (.\telegraf.exe --config-directory .\telegraf.d\ --test), copy the content in a textfile and count the lines (lines = metrics).
INTERVAL = The average interval that Telegraf collects the metrics.
Calculation: (METRIC_BUFFER / METRICS_PER_INTERVAL) * INTERVAL / 60 = Number of minutes Telegraf buffers metric locally
Example: (10000 / 700) * 30 / 60 = ~7 minutes