By the time the aggregator.run() was called about 600ms already passed since setting now which was skewing up the aggregation intervals and skipping metrics.
Adjust logic in functions responsible for passing metrics in order to be able
to process them correctly in case where pass and drop are defined together.
Looping the metrics back through the same channel could result in a
deadlock, by using a new channel and locking the processor we can ensure
that all stages can make continual progress.
It is not possible to encode a measurement, tag, or field whose last
character is a backslash due to it being an unescapable character.
Because the tight coupling between line protocol and the internal metric
model, prevent metrics like this from being created.
Measurements with a trailing slash are not allowed and the point will be
dropped. Tags and fields with a trailing a slash will be dropped from
the point.
main reasons behind this:
- make adding/removing tags cheap
- make adding/removing fields cheap
- make parsing cheaper
- make parse -> decorate -> write out bytes metric flow much faster
Refactor serializer to use byte buffer
in this commit:
- chunks out the http request body to avoid making very large
allocations.
- establishes a limit for the maximum http request body size that the
listener will accept.
- utilizes a pool of byte buffers to reduce GC pressure.
in this commit:
- centralize logging output handler.
- set global Info/Debug/Error log levels based on config file or flags.
- remove per-plugin debug arg handling.
- add a I!, D!, or E! to every log message.
- add configuration option to specify where to send logs.
closes#1786
started working on this with the idea of fixing #1623, although I
realized that this was actually just a documentation issue around
a toml eccentricity.
closes#1623
Piotr Popieluch
atzoum
Daniel Nelson
Maximilien Richer
DanKans
Seuf
Daniel Nelson
Patrick Hemmer
Daniel Nelson
tjmcs
Cameron Sparr
Alex Zorin