[BACKPORT] Send TCP metrics for bytes sent and received to server#497
[BACKPORT] Send TCP metrics for bytes sent and received to server#497mdumandag merged 2 commits intohazelcast:4.2.zfrom
Conversation
Codecov Report
@@ Coverage Diff @@
## 4.2.z #497 +/- ##
==========================================
- Coverage 95.16% 94.82% -0.35%
==========================================
Files 345 345
Lines 17552 17568 +16
==========================================
- Hits 16704 16659 -45
- Misses 848 909 +61
Continue to review full report at Codecov.
|
It would be good to send `tcp.bytesReceived` and `tcp.bytesSend` metrics in the metrics blob to the server, so that, someone can track how much traffic is passing between the server and the client. I have manually tested the stats exported to Prometheus and verified that these two new stats are displayed there correctly.
2956596 to
3c2babf
Compare
srknzl
left a comment
There was a problem hiding this comment.
LGTM with one minor comment but why codecov show coverage decreased a lot?
| @@ -1,10 +1,11 @@ | |||
| import time | |||
| import unittest | |||
There was a problem hiding this comment.
This import is not used anymore I think
There was a problem hiding this comment.
Also don't you check unused imports?
There was a problem hiding this comment.
Corrected it, thanks for the catch.
We use black to lint the code, but unfortunately, it does not have this feature. It seems autoflake and pylint have it, but enabling them requires a bit of work to get rid of their warnings. There was a task about this in Jira
It would be good to send
tcp.bytesReceivedandtcp.bytesSendmetrics in the metrics blob to the server, so that, someone can
track how much traffic is passing between the server and the client.
I have manually tested the stats exported to Prometheus and verified
that these two new stats are displayed there correctly.
Backport of #496