Hi Scott,
Thanks for responding.
As part of my test. I have also run my own program wherein I used both
random strings and json messages.
I was aware that snappy may not perform good on random strings. So I used a
10kb json message and sent 10,000 messages of that same json.
So with this test payl
Snappy and lz4 do not have entropy encoders (
https://en.wikipedia.org/wiki/Entropy_encoding). If your data is random
text they will not compress. If your text is a string of all zeros or any
repeating pattern, it will compress significantly. If its something like
JSON, or XML it will compress.
I have some updates on this.
I tried this on latest kafka 2.8. Ran my application. Results are same,
snappy and lz4 dont seem to be working as uncompressed and compressed
storage both measure the same.
*I even tried kafka-producer-perf-test tool*. Below are the results
Without any compression:
==
Hello,
There are times when Kafka would dump log messages at a rate such as 5
messages per second or 3 messages per second which are duplicated every
second. I have tried BurstFilter in log4j2 however it can swallow messages.
I am considering usage of Script filter of log4j2 but haven't tried
Hi Stein,
There is no intention to annoy anyone here. I just wanted know the reason for
the issue and want to resolve the issue.
The earlier email was marked as spam so I was not aware the my email received
response. After receiving this email, I checked spam folder and found the email
from Da
Laxmikant Shete,
You have already been told to stop spamming multiple lists across the
Apache Software Foundation. I am cc'ing them to let them know: any
further progress on this issue will be handled directly with the
Infrastructure team, which you can reference via us...@infra.apache.org. If
you
