Replies: 1 comment
-
simonswine can you please help on this |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
We can see below values in config. Do we have api to check number of profile received and size like uncompressed_bytes in distributor and ingester component.
# Maximum size of a profile in bytes. This is based off the uncompressed size.
# 0 to disable.
# CLI flag: -validation.max-profile-size-bytes
[max_profile_size_bytes: | default = 4194304]
# Maximum number of samples in a profile. 0 to disable.
# CLI flag: -validation.max-profile-stacktrace-samples
[max_profile_stacktrace_samples: | default = 16000]
For java we are running example simple program below within aks cluster
https://github.com/grafana/pyroscope/blob/main/examples/java/simple/Main.java
Observation -
We are able to scale by increasing the replica count. But we are not able to achieve expected throughput.
Increasing X replica for Go application increases the load X times but for java application increasing X replica we can't achieve X times load
Replicas Samples in billion/sec(approx)
1 1
5 4
10 8
50 15
100 29
Beta Was this translation helpful? Give feedback.
All reactions