Hi,
What are the recommended limits in terms of performance for:
metric_batch_size
metric_buffer_limit
I’m near to the limit configured below in my telegraf.conf file and for sure I’ll go > 100k, for this reason, I need to know the limitations and the recommendations if possible.
I have a telegraf.conf that is used for collecting events from some applications “node.js, spring boot 2.x” and I have these values now:
$ cat /etc/telegraf/telegraf.conf
[global_tags]
[agent]
interval = “16s”
round_interval = true
metric_batch_size = 90000
metric_buffer_limit = 100000
collection_jitter = “0s”
flush_interval = “16s”
flush_jitter = “0s”
precision = “”
debug = false
quiet = false
hostname = “”
omit_hostname = false[[outputs.prometheus_client]]
listen = “:9373”
path = “/metrics”
expiration_interval = “16s”
string_as_label = true
Thank you in advance.
Kind Regards,
