Replies: 2 comments
-
|
I believe that a proof-of-concept would be good to have anyway. Many questions so far — how do we integrate it with the CI, how will it affect packaging etc. But sounds interesting, and the results might be reused in other parts of the project. |
Beta Was this translation helpful? Give feedback.
-
|
So the rewrite of the exporter is already almost done ... I am about to delve deeper ... Alright. My cogitations are similar, could be great, but also mingling many notions. What if nimlang is used too much within pyroute2, is it then nimroute2 under the surface. :D ;) That was "semi-serios" ... My stance is, for all the "low" layer netlink interactions python shines with its fantastically amazing prototyping nature. But at all the post processing, maybe even at the marshalling of the netlink results, it does not necessarily so. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi,
Short:
would nimlang code sections be accepted in pyroute2 - e.g. for ss2 utility?
Long:
I'm planning to enhance the scalability and performance of - prometheus_ss_exporter
Currently even for small flow counts, the current python ss2 and exorter logic reaching quickly its limits - running into inacceptable scrape times.
The pyroute2/sockdiag is an acceptable overhead as such. I'm also looking into rust netlink implementations for sockdiag, neli or alike, but these appear to be very immature in comparison as of writing.
Would that imply the ss2 has to wander out the project? Other implications or thoughts?
Kindly let me known.
Thanks in advance
Beta Was this translation helpful? Give feedback.
All reactions