Dear Users
I am collecting data from several instruments that have data in three components (x,y,z), each one second. The instrument outputs are in CSV format, and I am injecting data into influxdb using telegraf. I’ve read a lot of documentation about the best combination of parameters to reduce “Metric buffer overflow” type messages, but no combination of parameters seems to have any effect… I’m a bit confused at this point, and I’d like to know if there’s any other way to define the parameters to at least reduce the messages from telegraf… any ideas?
telegraf-1.32.3-1
influxdb-1.11.8
my telegraf config is:
[agent]
interval = “5s”
round_interval = true
metric_batch_size = 25000
metric_buffer_limit = 400000
collection_jitter = “0s”
flush_interval = “5s”
flush_jitter = “2s”
precision = “1s”
debug = true
[[outputs.influxdb]]
urls = [“http://127.0.0.1:8086”]
database = “gn”
skip_database_creation = true
retention_policy = “oneday”
timeout = “5s”
username = “rod”
password = “xxxxxx”
write_consistency = “any”
messages from the system:
telegraf[2993954]: D! [outputs.influxdb] Wrote batch of 25000 metrics in 612.824738ms
telegraf[2993954]: W! [outputs.influxdb] Metric buffer overflow; 88286 metrics have been dropped
telegraf[2993954]: D! [outputs.influxdb] Wrote batch of 25000 metrics in 683.299503ms
telegraf[2993954]: W! [outputs.influxdb] Metric buffer overflow; 38221 metrics have been dropped
